Sage Journals: Discover world-class research

Abstract

When treating diabetic cats, the primary aim is to control clinical signs without causing clinical hypoglycaemia. Secondary goals are to maximise the chances of attaining diabetic remission and to minimise the risk of complications due to chronic hyperglycaemia. A treatment plan that is convenient for the owner is important for compliance. Underweight or overweight diabetic cats should be fed with the aim of normalising bodyweight. Current evidence suggests that non-obese diabetic cats can be fed ad libitum. The oral hypoglycaemic drug glipizide is well established as a treatment for about a third of diabetic cats, which have residual beta cell function. Preliminary studies on other oral agents such as vanadium salts, metformin, and troglitazone indicate a potential use in some diabetic cats. Insulin treatment remains the treatment of choice for the majority of diabetic cats. Choice of insulin, dose rates and monitoring of treatment are discussed.

This is the second of a two-part review on diabetes in cats. Part I (Rand 1999) discussed the pathogenesis and diagnosis of diabetes in cats. Knowledge of pathogenesis is important for understanding treatment issues.

Sick ketotic diabetic cats

Diabetic cats that present with diabetic ketosis or ketoacidosis (DKA) may be severely ill, and their treatment is a medical emergency. In two separate studies, each of 104 diabetic cats, there were 12 cats (11.5%) and 38 cats (36.5%) diagnosed with DKA on the basis of ketosis and metabolic acidosis (Crenshaw & Peterson 1996, Goossens et al 1998). In Goossens' study, eight of 12 cats with DKA died or were euthanased in the initial hospitalisation period (Goossens et al 1998). Acute necrotising pancreatitis was a frequent concomitant problem in these cats. The treatment of DKA involves addressing underlying disease processes present, correcting fluid, electrolyte, and acid-base imbalances, and lowering blood glucose (Macintire 1995).

Depressed, dehydrated, hypothermic, or pyrexic cats require hospitalisation. In dehydrated diabetic cats, fluids are as important as insulin therapy. For most dehydrated diabetic cats, the type of fluid is not as important as the fluid itself, and a balanced electrolyte solution, preferably with lactate, is a good initial choice. Generally, fluid administration rates of 100 to 150 ml/kg/24 h are appropriate for dehydrated diabetic cats. For severely hyperglycaemic (>32 mmol/l) dehydrated cats, half strength sodium chloride (0.45% NaCl) may be the preferred choice of fluid for the first few hours, because these cats are markedly hyperosmotic. However, it should be used with careful monitoring of blood glucose and electrolytes. Bicarbonate supplementation is rarely indicated because acidosis improves rapidly with fluid and insulin therapy.

Ketoacidotic cats may initially be hyperkalaemic. Acidosis causes intracellular potassium to be exchanged with extracellular hydrogen ions, which may result in hyperkalaemia. Extracellular potassium concentration falls quickly as potassium moves intracellularly with insulin and fluid therapy, and as acidosis is corrected. Whole body potassium depletion occurs when glucose and fluid induced diuresis causes renal potassium loss (Dow et al 1987, Feldman & Nelson 1996). Hypokalaemia may occur within 12–24 h in cats that were hyperkalaemic on admission. Hypokalaemic myopathy and weakness can occur with plasma potassium concentrations less than 3.5 mmol/l, and levels below 2 mmol/l should be considered an emergency because of the possibility of death from respiratory paralysis (Dow et al 1987). Potassium therapy should commence immediately in normokalaemic acidotic cats, and within 12–24 h after commencement of fluid therapy in initially hyperkalaemic cats. If possible, potassium therapy should consist of both oral and intravenous therapy. Potassium chloride should be added to fluids at between 40 and 80 mmol/l, and potassium gluconate given at 2–3 mmol/cat per os, one to three times per day. Potassium concentrations need to be monitored carefully in the first few days of fluid therapy to avoid hypo- or hyperkalaemia.

Often ketoacidotic cats become hypophosphotaemic within 12–24 h after commencement of insulin therapy (Feldman & Nelson 1996). Metabolic acidosis causes cellular and whole body loss of phosphorus, similar to potassium (Willard et al 1987). Signs of hypophosphataemia (<0.5 mmol/l) in cats are haemolysis, muscle weakness, as well as neurological problems such as seizures and ataxia (Forrester & Moreland 1989). Phosphorus can be given subcutaneously at 0.14–1.26 ml/kg. Alternatively, potassium and phosphorus can be supplemented in the intravenously administered fluid by calculating the potassium requirement and supplying it as 50% potassium chloride and 50% potassium dihydrophosphate (Feldman & Nelson 1996).

In dehydrated diabetic cats, insulin is best given intramuscularly or intravenously, and blood glucose levels monitored carefully. The aim of insulin therapy is to decrease blood glucose levels slowly, while decreasing ketoacidosis. Problems can be encountered when blood glucose levels fall too rapidly, as diabetic cats have adapted to prolonged hyperosmolality and hyperglycaemia. Aim for a glucose drop of 3–4 mmol/l/h and do not exceed 6 mmol/l/h. In dehydrated, depressed ketoacidotic cats, regular insulin can be administered intramuscularly at an initial dose of 0.2 iu/kg, and continued hourly at 0.1 iu/kg. To achieve blood glucose decreases of between 3 and 4 mmol/l/h, insulin doses may have to be increased to 0.2 iu/kg/h or decreased to 0.05 iu/kg/h (Feldman & Nelson 1996). Hourly injections should be continued until blood glucose reaches 16 mmol/l, when injections of 0.1–0.4 iu/kg should be given every 4–6 h to achieve a final blood glucose of between 11 and 14 mmol/l. If blood glucose levels fall below 8 mmol/l in the first 36 h, a 50% dextrose (glucose) solution should be added to the intravenous fluids to produce a 5% dextrose solution. Alternatively, regular insulin can be given intravenously at 0.05–0.1 iu/kg/h and adjusted to achieve a glucose decrease of 3–4 mmol/l/h. If the cat is very dehydrated and ketotic and regular insulin is not available for several hours, then other longer-acting forms can be given intramuscularly. Lower the dosing frequency to every 4–6 h if using either lente or ultralente forms (0.3–0.5 iu/kg), and monitor glucose very carefully, especially after the second injection. Subcutaneous injection of maintenance insulin, such as lente, PZI or ultralente, can commence when the cat is well hydrated.

For further details on treating DKA, and other severe forms of diabetes mellitus in cats, the reader is referred to reviews on this specific disorder and its clinical management (Macintire 1995, Nicholls & Crenshaw 1995a, 1995b, Bruskiewicz et al 1997).

Therapeutic goals for otherwise healthy diabetic cats

The principal goals of treating diabetes in cats are to correct the major clinical signs of diabetes (weight loss, polydipsia/polyuria, and polyphagia or inappetence), and to prevent diabetic ketoacidosis. Additional goals are to minimise the risk of hypoglycaemia by appropriate dose adjustment of insulin or oral hypoglycaemic agents. Appropriate therapy of cats may also lead to diabetic remission (Nelson et al 1992a), although currently it is not possible at diagnosis to predict which cats are more likely to go into remission. Treatment that is most effective in correcting hyperglycaemia is more likely to lead to remission by facilitating the recovery of the pancreatic beta-cells from glucose toxicity.

The long-term complications of diabetes in cats are of less concern than with other species, and hindlimb neuropathy is the most significant problem (Munana 1995). Tight glycaemic control is pursued in human diabetics to avoid the long term complications associated with chronic hyperglycaemia such as retinopathy, but this leads to an increase in the frequency of clinical hypoglycaemic episodes (Reichard 1994). Because of the risk of hypoglycaemia in insulin-treated diabetic cats (Whitley et al 1997) and the lack of life-threatening complications associated with moderate hyperglycaemia, pursuing tight glycaemic control is not advised.

Treatment modalities

Diet

General considerations . Unlike humans and dogs, normal cats fed ad libitum will eat 12–20 meals spread throughout the day and night, and traditionally the major energy source is protein rather than carbohydrate (Morris & Rogers 1983, Houpt 1991). Another important species difference is that post-prandial hyperglycaemia does not occur in cats fed typical cat foods (Kienzle 1994, Martin & Rand 1997). Therefore there does not seem any need to match the timing of the insulin dose to meals, which is important in the control of blood glucose in human diabetics (Eaton et al 1978). Diabetic cats can be fed ad libitum and allowed to follow a normal feeding pattern of multiple small meals (Kane 1989, Martin & Rand 1997). Feeding multiple kilojoule-restricted meals may also be advantageous for weight loss in obese diabetic cats (Hand et al 1989).

Diabetics have energy loss via glucose in the urine, and may have maldigestion/malabsorption due to previous pancreatitis or diabetic enteropathy (Diehl 1995), therefore maintenance energy requirements may be higher than in normal cats. Polyphagic diabetic cats should not have their food intake restricted unless as part of a planned weight loss program for obesity.

High fibre diets are recommended in diabetic humans to blunt post-prandial hyperglycaemia and increase peripheral insulin sensitivity (Anderson & Akanji 1991). A high fibre diet (12% insoluble fibre) was found to be beneficial in some diabetic cats, in comparison to a very low fibre diet (1%) (Nelson et al 1994). However, varying carbohydrate contents of the diets confounded the results of this study, as the low fibre diet had substantially more carbohydrate. There are no studies in diabetic cats showing an advantage of a high-fibre diet over a good quality commercial cat food with a moderate fibre content (eg, 5%). The benefit of increased fibre, isolated from changes in dietary carbohydrate, may be less in cats than in humans because of the absence of post-prandial hyperglycaemia in cats.

Although only foods containing substantial amounts (36.1% w/w) of glucose (and not other carbohydrates) led to post-prandial hyperglycaemia in a study of normal cats (Kienzle 1994), for diabetic cats it is probably advisable to avoid diets that are unusually high in carbohydrate, and to avoid glucose-containing ‘treats’.

Feeding recommendations for diabetic cats . Diabetic cats, particularly soon after diagnosis, may have reduced or no appetite. This may require the initiation of assisted feeding, such as hand feeding of highly aromatic, palatable food (Laflamme et al 1993). Once eating normally, diabetic cats with a normal bodyweight should be fed a good quality feline diet ad libitum, and have their bodyweight monitored. As the diabetes is stabilised with therapy, bodyweight often increases. If the cat starts to become overweight, it should be changed to a moderately calorie-restricted diet, such as a geriatric diet, or a diet for less active cats. Most high fibre diets are calorie-restricted, and these diets may be unsatisfactory for some non-obese diabetic cats. Some normal weight diabetic cats lose weight, and underweight cats may fail to achieve normal bodyweight, despite polyphagia when fed this type of diet. The increased stool volume may also be unacceptable to the owner.

When diabetic cats are underweight, energy dense foods such as conditioning or growth diets should be used. As weight approaches normal, cats can be switched to maintenance cat foods. The best combination of protein, fat, and carbohydrate for diabetic cats is currently unknown.

Obese diabetics . Weight loss and attainment of normal bodyweight is a treatment goal of overweight diabetic cats, as obesity in cats reversibly decreases glucose tolerance (Nelson et al 1990, Biourge et al 1997). Calorie restriction should be conservative, and matched to the individual cat by ongoing monitoring of food intake and body-weight (Sloth 1992). Cats should be fed approximately 70% of their maintenance energy requirements for their target weights (Sloth 1992). Severe calorie restriction or starvation should not be used for weight loss because diabetic cats have an increased risk of hepatic lipidosis (Bruskiewicz et al 1997). It is important to ensure that the cat will eat the diet when a change is made to a weight reducing diet. To avoid excessively rapid weight loss, slow introduction of a diet over 2 weeks may be more successful in getting the cat to eat a new diet than abrupt dietary changes. Once the estimated target weight is reached, re-evaluation of the cat's body condition should be made.

Oral hypoglycaemics

Approximately one-third of diabetic cats achieve good clinical control with oral hypoglycaemic drugs. However, in the majority of cats, the control of diabetes is not as good with oral hypoglycaemics as with insulin, and in some cats it may be more difficult for the owners to administer oral medication than insulin injections. When owners are unwilling to use insulin, oral agents may be lifesaving for a cat that would otherwise be euthanased. Insulin treatment can be started if oral treatment is unsuccessful, which is usually evident within 4–6 weeks. By this time, owners may be more amenable to the use of insulin (Feldman et al 1997). Oral hypoglycaemic agents are generally not recommended in sick ketotic or anorexic diabetic cats, which should be treated with insulin (Nelson et al 1993). Treatment for these cats may be subsequently attempted using oral drugs, once their condition has improved with insulin and supportive care.

The oral sulphonylurea hypoglycaemic drug glipizide has been used successfully in the treatment of diabetic cats (Miller et al 1992, Nelson et al 1993, Feldman et al 1997). The usual dose of glipizide is 5 mg bid, regardless of the body-weight of the cat, though lower doses (1.25–2.5 mg bid) may initially be used to acclimatise the cat to the drug and reduce side-effects (Miller et al 1992, Nelson et al 1993, Feldman et al 1997). Glipizide requires less intensive clinical monitoring than insulin treatment, due to a lower risk of clinical hypoglycaemia, and may create sufficient glycaemic control to lead to diabetic remission (Nelson et al 1993, Feldman et al 1997). Hypoglycaemia may still occur in cats given glipizide, particularly in those cats whose hyperglycaemia is mild (12–15 mmol/l), or those which are attaining diabetic remission (Nelson et al 1993, Feldman et al 1997).

Glipizide stimulates beta cells to produce insulin and is only effective in diabetics which have functional beta cells (Zimmerman 1997), which were about 40% of diabetic cats reported in a recent study (Feldman et al 1997). Good control has been reported in about 35–40% of diabetic cats treated with glipizide (Feldman et al 1997), although the success rate depends on the population of cats studied and criteria for good control. As glipizide may cause transient elevation of liver enzymes and in some cases icterus, monitoring hepatic enzymes is recommended during glipizide treatment, which adds to treatment costs (Miller et al 1992, Nelson et al 1993, Feldman et al 1997). Oral hypoglycaemic drugs fail to function adequately in some human patients over time, as the number of functional beta cells decreases (Chow et al 1995), and this may also occur in some cats on glipizide treatment (Feldman et al 1997). Of concern with drugs which stimulate insulin and amylin secretion, such as glipizide, is that amyloid deposition may be accelerated and lead to further loss of beta cells (Hoenig & O'Brien 1998, Rachman et al 1998). This may lead to insulin-dependence and poorer control of blood glucose than if other types of drugs, including insulin, had been used initially.

There are a variety of other oral agents that are used in humans for the treatment of diabetes. Preliminary studies in cats have been performed with vanadium, which increases insulin receptor sensitivity. However, as a sole agent it was only useful early in the pathogenesis of diabetes when blood glucose was relatively low (Greco 1997). By the time a diagnosis of diabetes is made, vanadium is unlikely to be effective by itself. A further preliminary report of a comparative study using PZI insulin with and without concurrent vanadium administration (45 mg orally once daily), showed that vanadium may reduce the required dosage of insulin (Fondacaro et al 1999). However, the biologically available forms of vanadium, such as di- and tripicolinate or bismaltolatooxo, may not be readily obtainable by practitioners.

Chromium, a compound next to vanadium on the periodic table and with similar properties, has been shown to have small but significant effects on blood glucose in some normal cats, although there are no reports of its efficacy in diabetic cats (Appleton et al 1999). It has a wide safety margin and if useful, it could be used as an adjunctive dietary additive for diabetic cats. The biologically available tripicolinate form is usually used.

There are anecdotal reports of the use in diabetic cats of acarbose, an agent which blocks glucose absorption from the gastrointestinal tract (Greco 1999), but there are no studies documenting its effectiveness. The lack of post-prandial hypoglycaemia in cats may reduce the usefulness of acarbose in diabetic cats, but it may have a role in combination with other agents (Greco 1999).

One drug which has potential to be useful in the treatment of feline diabetes is metformin, which enhances peripheral insulin sensitivity and reduces hepatic glucose output (Schernthaner 1985). To date, the only study reported describes the pharmacokinetics of metformin in normal cats (Michels et al 1999). Studies on metformin in diabetic cats are currently in progress.

The pharmacokinetics of the insulin-resistance reducing drug troglitazone have been reported for normal cats in a recent study (Boudinot et al 1999). An anecdotal report of its use in diabetic cats at a dose of 200 mg once daily suggests its efficacy may be limited (Greco 1999).

In summary, glipizide is the only oral hypoglycaemic drug that has well documented effectiveness in diabetic cats. A preliminary study suggests vanadium may be a useful adjunct to insulin treatment to improve insulin sensitivity.

Insulin

The mainstay of treatment in diabetic cats is injectable insulin. The types of insulin available and the rationale for their use in cats has been the subject of several papers and reviews (Moise & Riemers 1983, Schaer 1983, McMillan & Feldman 1986, Nelson et al 1992b, Bertoy et al 1995, Broussard & Wallace 1995, Dowling 1995, Goossens et al 1998) (see Table 1). The following discussion concentrates on some aspects of insulin treatment, treatment monitoring, and making insulin dose rate adjustments.

Table 1.

Examples of insulins for diabetic cats

Insulin	Advantages	Disadvantages	Trade names
Long acting
PZI	Can be given once-daily in many cats	May have an unpredictable onset and duration of action. 24-h blood glucose curves should be performed to ensure nocturnal hypoglycaemia is not occurring. Accurate dosing with 100 iu/ml preparation may be a problem. Prolonged hypoglycaemia may occur	PZI insulin, 40 iu/ml* Insuvet, 100 iu/ml (Europe)^†
Human or porcine ultralente	Can be used successfully once-daily in some cats	As for PZI, but fewer cats can be managed on once daily ultralente than PZI. The majority of cats will require twice daily dosing. Some cats will have better control with lente insulin. Accurate dosing with 100 iu/ml preparation may be a problem	Ultratard or Ultratard MC^† (porcine) 100 iu/ml Humulin UL** 100 iu/ml
Bovine ultralente	Can be given once-daily in many cats	As for PZI, but may be so poorly absorbed in some cats that very high doses are needed to adequately lower blood glucose. Accurate dosing with 100 iu/ml preparation may be a problem. Availability is a problem in many countries
Intermediate acting NPH (human or porcine)	More predictable onset and duration of action than long-acting insulins. 12-h blood glucose curves taken during the day should give an indication of blood glucose status during the night as well as the day	Requires twice-daily dosage. In many cats, duration of action may be too short (<3 h). Accurate dosing with 100 iu/ml preparation may be a problem	Protophane^† Humulin N** (human) 100 iu.ml
Human lente	Similar to NPH above	Requires twice-daily dosing. Accurate dosing with 100 iu/ml preparation may be a problem	Monotard HM^† Humulin L** (human) 100 iu/ml
40 iu/ml porcine lente	As for lente, but the veterinary 40 iu/ml preparation allows more accurate dosing of cats. Predictable onset and duration of action in most cats	Requires twice-daily dosing. Rarely has too short a duration of action	Caninsulin^††
Short acting Soluble insulin	Suitable for intravenous infusion	Not suitable for long-term management of diabetic cats due to the short duration of action	Actrapid^† Humulin R** (human) 100 iu/ml

*Blue Ridge Pharmaceuticals (USA);

†

Schering-Plough Animal Health;

‡

Novo Nordisk;

Aza/Eli Lilly;

††

Intervet (The Netherlands).

Choice of insulin . The choice of insulin for diabetic cats centres around minimising the number of injections required each day, while still achieving adequate diabetic control. Insulins used for maintenance are those classified as either intermediate duration (lente or NPH [Isophane]), or long acting (ultralente or protamine zinc insulin [PZI]) (Greco et al 1995). The species of origin of the insulin effects the duration of insulin action, with beef insulin being longer acting than pork, and human insulin the shortest in duration (Brange et al 1990). Despite the variation in insulin types used, no correlation was found between the type of insulin and diabetic control (Goossens et al 1998).

The antigenicity of an injected insulin is proportional to the difference in its amino acid sequence to the amino acid sequence of the animal's native insulin (Neubauer & Schone 1978, Schernthaner 1993). Cat insulin is most similar to beef insulin (one amino acid difference), less so to pork (three amino acids difference), and least similar to human (four amino acids difference) (Hallden et al 1986). Unfortunately in many countries, the withdrawal of animal insulins for human use, and insulin containing the amino acid protamine (PZI), has decreased the range of suitable insulins for the treatment of diabetes in cats.

Human insulin has been found to be effective in the treatment of feline diabetics (Nelson et al 1992b, Bertoy et al 1995). Although most different from feline insulin, it has not been found to be associated with substantial antibody production. Human insulins are sold as 100 iu/ml concentration or greater in most countries, which may cause difficulties in dosing cats, as doses of less than 2 iu cannot be measured accurately in a U-100 syringe (Casella et al 1993). Dilution of insulin may significantly shorten its duration of action (Binder 1969, Chantelau et al 1985, Heinemann et al 1992).

The long-acting insulin PZI insulin was particularly favoured for the treatment of cats until it became unavailable as a human product, forcing a change to shorter-acting insulins, which have subsequently been found to give as good, and in some cats better glycaemic control. PZI insulin is now available as a veterinary product in the USA as a 40 iu/ml 90% beef–10% pork suspension (PZI insulin; Blue Ridge Pharmacy) and in the UK as a 100 iu/ml beef suspension (Insuvet Protamine Zinc, Schering-Plough Animal Health). PZI can be used once daily in 50% of cats (Moise & Riemers 1983, Goossens et al 1998), but is noted for its unpredictable onset and duration of action (Moise & Riemers 1983). Prolonged marked hypoglycaemia may occur in some cats on PZI and, in some cats, glycemic control is poor, possibly because of inadequate absorption. PZI is particularly useful for cats in which the duration of action of lente insulin is too short to give adequate glycaemic control, and when owners are unwilling to give twice-daily injections. Many cats have good to excellent clinical control with PZI, but cats which do not achieve good control should be tried on a shorter acting insulin such as lente, before the poor control is attributed to insulin resistance.

Human ultralente insulin can be used once daily in many cats, but the majority of cats (92%) will require twice daily treatment (Goossens et al 1998). It is readily available through human pharmacies, although not licensed for veterinary use. Because of this dosing frequency and 100 iu/ml concentration, it does not offer an advantage over lente or PZI insulin, especially if either of these are available at 40 iu/ml. Bovine ultralente insulin may also be used, and because of its relatively peakless absorption, in some cats it potentially provides better glycaemic control than the shorter-acting insulins. However, in some cats blood glucose may not be significantly lowered, apparently because of poor absorption (Broussard & Peterson 1994). Bovine ultralente is no longer available as a human product in many countries, including Australia.

In some cats, blood glucose does not fall significantly and control of clinical signs is poor when using the longer-acting insulins (human or bovine ultralente, or PZI), even when doses of 1 iu/kg or greater are being used. A change to a better absorbed, shorter acting insulin (lente or NPH) may improve diabetic control in these cats (Bertoy et al 1995).

The lente insulins have a more predictable onset and duration of effect in cats than ultralente and PZI (Martin & Rand, unpublished data), and have been recommended as a first choice of insulin (Rand 1997) or where ultralente is failing to achieve adequate control (Bertoy et al 1995). Porcine lente insulin is available in many countries as a proprietary veterinary product, Caninsulin (Intervet International BV, The Netherlands), at a concentration of 40 iu/ml, which is preferable to 100 iu/ml for use in cats. Porcine lente insulin needs to be given twice daily in all diabetic cats (Martin & Rand, unpublished data). Caninsulin provides good to excellent clinical control in the majority of cats, although the occasional cat has a very early peak of action (3 h or less), and has better control with a longer-acting insulin (Martin & Rand, unpublished data). NPH insulin has a similar but slightly shorter duration of action than lente insulin, and usually is given twice daily (Moise & Riemers 1983). Human NPH is readily available through most pharmacies and provides adequate clinical control in a substantial proportion of cats. It is less useful in cats than 40 iu/ml porcine lente (Caninsulin) because it is only available at 100 iu/ml concentration and has a shorter duration of action. In cats with inadequate control with PZI or ultralente insulin, human or bovine NPH and lente insulins are useful alternatives when 40 iu/ml porcine lente insulin is unavailable. The imminent release of new insulin pens with a 1 iu minimal total dose and incremental doses of 0.5 iu may lead to wider use of human NPH insulin. Lente, ultralente and PZI are currently unsuitable for use in insulin pens.

In summary, good to excellent clinical control and adequate glycaemic control can be achieved with NPH, lente, ultralente or PZI insulin in the majority of diabetic cats. The choice of insulin needs to be based on what is available, and the suitability for the individual cat and convenience for the owner. Concentration is an important consideration and a 40 iu/ml insulin is more suitable than 100 iu/ml for administration of the small doses of insulin that diabetic cats require. Because of legal implications and product support, use of a licensed veterinary product rather than ‘off label’ use of a human insulin is recommended. In general, the blood glucose response with NPH or lente insulin is more predictable than with PZI. Up to one in five cats on PZI or ultralente have inadequate control, in part from poor absorption, and in a few cats, the duration of action of lente insulin may be too short to achieve adequate glycaemic control. Inadequate duration of action is more often a problem with NPH, but can be overcome with administration three times a day. Approximately 50% of cats on PZI can be controlled with once-daily insulin, whereas up to 90% of cats with ultralente and all cats on lente insulin need twice-daily dosing.

Starting dose of insulin . The type of insulin chosen and the baseline blood glucose of the cat determine the initial dose of insulin. Insulins which have a longer duration of action do not tend to lower the blood glucose as much as the same dose of a shorter acting insulin (Nelson et al 1992b). Our current regime is to start cats which have marked hyperglycaemia (blood glucose ≥20 mmol/l) on 0.5 iu/kg insulin when using porcine lente, and to start cats with less pronounced hyperglycaemia (blood glucose <20 mmol/l) on 0.25 iu/kg insulin. If the blood glucose response cannot be measured, then it is best to start with a minimal total dose of insulin (1–2 iu). The evaluation of the response to insulin using serial blood glucose determinations and the rationale for insulin dose rate changes is discussed below.

Inter-day variability of the insulin response . The responsiveness of diabetic cats to insulin can be frustratingly variable, both from day to day and over longer periods. For example, we have seen differences in the nadir (lowest) blood glucose of greater than 5 mmol/l on consecutive days. The reasons for inter-day variability may come from variations in the amount of insulin administered, in the concentration of insulin per batch, in the absorption of insulin, the availability of insulin in the plasma to the insulin receptors, and the magnitude of the responsiveness of the peripheral tissues (peripheral insulin sensitivity). Because of inter-day variation in insulin effects, the serial blood glucose results should be interpreted conservatively for the purpose of making dose rate adjustments.

The rate of insulin absorption may be affected by the type of insulin, its species of origin, concentration, dose, and handling (Galloway et al 1981, Hildebrandt 1991, Home & Alberti 1992). Practitioners should re-evaluate the blood glucose response if the insulin type is changed, the insulin dose is changed, or the insulin concentration is changed. Although U-100 and U-40 have a similar time action profile of blood glucose in humans (Chantelau et al 1985, Heinemann et al 1992), dilutions of insulin by a factor of five or greater may significantly shorten insulin absorption and the blood glucose response (Binder 1969, Chantelau et al 1985, Heinemann et al 1992). Clients should be advised carefully on the appropriate methods of storing, handling, and drawing up of insulin (Greco et al 1995). If U-100 syringes are being used with a U-40 insulin, it is important to ensure that the correct dose rate is calculated.

Variability in insulin absorption is suspected to account for some of the day-to-day variability in the response to a consistent insulin dose in humans; factors include: blood flow at the injection site, anatomical location of the injection site, circulating epinephrine, ketoacidosis, amount of fat at the injection site, exercise or local massage, injection technique (subcutaneous versus intramuscular), ambient temperature, scarring and lipodystrophy at the injection site, and/or anti-insulin antibodies causing local degradation of the insulin at the injection site (Galloway et al 1981, Binder et al 1984, Skyler 1988). The variability in insulin absorption has not been studied in diabetic cats, but based on the findings in humans it is advisable to use a consistent injection site and to minimise stress at the time of injection to avoid circulating epinephrine release.

Day-to-day variation in insulin sensitivity is thought to be a significant cause of inter-day variation in normal (Steil et al 1994) and diabetic (Ziel et al 1988) humans. Studies in normal cats have shown inter-day variation in insulin sensitivity to be marked (Feldhahn et al 1998), and this may also be the case in diabetic cats.

Longer term variability in insulin response . In humans, the clinical term ‘brittle diabetes’ is used to refer to patients whose blood glucose responds in an unpredictable fashion to insulin injections, particularly where hypoglycaemia occurs (Schade & Burge 1995). One specific cause of brittle diabetes in humans is the loss of the normal counter-regulatory glucagon response to hypoglycaemia, which may lead to hypoglycaemia or post-hypoglycaemic hyperglycaemia through activated counter-regulation (the Somogyi effect) (Gerich 1988). Cats may exhibit the features of brittle diabetes, and the Somogyi phenomenon has been previously reported in insulin-treated diabetic cats (McMillan & Feldman 1986). Although the existence of the Somogyi phenomenon has been subsequently questioned in human diabetology (Kidson 1993), a reduction in the insulin dose may improve control in cats in which either hypoglycaemia and/or hyperglycaemia is recurrent. Hypoglycaemia is more common in dogs when intermediate acting insulin is given once daily, and in humans, when long-acting insulin is given twice daily. If glycaemic control is poor in an individual cat, especially if insulin seems to have little effect when previously it caused substantial lowering of glucose, it is safer to first try lowering the dose of insulin to 0.3–0.5 iu/kg (lente insulin) for 2–3 weeks and see if blood glucose or water intake improve. Poor control and apparent insulin resistance with persistent hyperglycaemia may result from excessive insulin dose.

As the hyperglycaemia of diabetes is reduced by insulin treatment over time, the insulin requirement of the cat may decrease (Garvey et al 1985, Link & Rand 1996). This effect requires 2 or more weeks to occur, hence a period of 2–4 weeks between insulin dose increases is recommended (Garvey et al 1985, Nijs et al 1989, Link & Rand 1996). This effect is probably due to the reversal of one or more of the three major metabolic lesions of type 2 diabetes: glucose toxicity, decreased insulin sensitivity, and increased hepatic glucose output, as hyper glycaemia is diminished (Garvey et al 1985). The insulin requirement may increase in some cats after several months of therapy as further loss of beta cells occurs.

In cats where some functional islet cells remain, resumption of beta-cell function may occur after at least 1–2 weeks of good glycaemic control, which may lead to cessation of the diabetes within 1–3 months (Link & Rand 1996). If the loss of beta cells is too great for diabetic remission to occur, the cat may remain insulin-dependent but with reduced insulin doses being required. Sudden or gradual reductions of more than 50% in the required insulin dose may occur even after several months of insulin therapy in cats, possibly through this mechanism (Martin & Rand, unpublished data).

A recent study found that diabetic control was not correlated with anti-insulin antibody titre in diabetic cats (Harb-Hauser et al 1998). Further studies are needed to determine if insulin antibodies affect diabetic control in certain feline diabetics with, for example, high insulin requirements or repeated episodes of hypoglycaemia.

Serial blood glucose curves . Serial blood glucose curves are essential for establishing and monitoring the blood glucose response to insulin, and for making insulin dose rate adjustments (Miller 1995, Zerbe 1999). When a low blood glucose (<3.0 mmol/l) occurs, samples should be taken every 30–60 min until the blood glucose rises, to determine if glucose should be administered. Typically, blood glucose is measured every 2 h, from immediately prior to one insulin dose to immediately prior to the subsequent insulin dose. If insulin is given once daily, ideally the blood glucose should be followed for 24 h, as nocturnal hypoglycaemia may occur even if the blood glucose is normal or elevated during the day. If the insulin is given twice daily, 12 h should be suitable, although it is uncertain if the overnight blood glucose curve will exactly match the daytime curve. When Caninsulin was used, the median nadir was found to be statistically higher during the night than during the day (Martin & Rand, unpublished data).

Pocket glucometers are sufficiently accurate to use to perform serial blood glucose curves in a clinical setting (Link et al 1997). The limitations of such meters, such as the maximum blood glucose measurable and the tendency to be falsely low at the lower end of the range, should be considered when interpreting results (Link et al 1997).

Stress resulting in struggling can acutely elevate the blood glucose by up to 10 mmol/l in normal cats (Kinnaird et al 1998), and possibly more in diabetic cats. This may lead to the false impression that the insulin dose is inadequate. If the insulin dose is increased as a consequence, hypoglycaemia may occur. If the cat is stressed or struggles before or during blood collection, results should be ignored and the procedure repeated later. Some cats may need to be hospitalised for 12–24 h before a serial blood glucose curve to allow them to settle into a foreign environment, however, hospitalisation may increase stress in other cats.

Blood collection . Blood collection can be performed by direct venepuncture, or peripheral or central venous catheterisation. Venepuncture is only suitable for the collection of a limited number of samples in diabetics cats, as bruising and intolerance of the procedure cause difficulties after a limited number of samples are taken in most cats. To provide venous access in a repeatable and non-stressful manner, placement of venous catheters may be necessary. Peripheral catheters can be placed in the cephalic or lateral saphenous veins, flushed with 20 iu/ml heparinised saline, and capped. These catheters can then provide access for small blood samples suitable for a pocket type glucometer for repeated sampling of up to 24 h or more in some cats, although not all cats will tolerate them (Burrows 1973, Link et al 1997). When blood glucose needs to be monitored over several days, or when peripheral vein catheters are poorly tolerated, jugular catheters are advised (Martin & Rand 1999b). New vacuum-assisted (Wess & Reusch 1999) or laser lancing devices may provide access to capillary blood without the need to obtain venous samples, and with less risk of stress to the cat. Non-invasive transdermal methods for measuring blood glucose are under development, and would be of great benefit in monitoring feline diabetics (Arnold 1996).

Insulin dose adjustments . From the serial blood glucose curve, several parameters may be recorded or calculated, including mean blood glucose, blood glucose at baseline (immediately before insulin), nadir (minimum) blood glucose, and the time to the nadir of blood glucose. Mean blood glucose gives an indication of overall glycaemic control, and the difference between the baseline and nadir gives an indication of the cat's responsiveness to the insulin. The time to the nadir gives an indication of how rapidly the cat is responding to the insulin injection. The nadir of blood glucose determines whether the insulin dose should be increased or decreased.

In practice the most important value to determine is the nadir blood glucose, as this indicates the maximum dose of insulin which can be tolerated by the cat. The dose should not be increased to the point where hypoglycaemic episodes are likely (<4 mmol/l). Given the potential for the response to insulin to be variable from day to day, and the possibility of a gradually improving insulin response, insulin dose increases should be made conservatively and infrequently (no more than every 2 weeks).

A suggested regimen is to re-evaluate the blood glucose response every 2–4 weeks (or immediately if clinical hypoglycaemia occurs) until good control is attained. In the early stages of treatment, a nadir of 5.0–9.0 mmol/l is a suitable target. If the nadir of blood glucose is above this level, the insulin dose is increased by 1 iu total dose. If clinical hypoglycaemia occurs, the insulin dose should immediately be decreased by 50%, and a serial blood glucose curve performed for the new dose (Greco et al 1995). Only in cats that show a very consistent response to insulin, and have good glycaemic control (mean blood glucose ≤12 mmol/l) with no hypoglycaemic episodes, can a nadir of 4.0–5.0 mmol/l be pursued.

When using a long acting insulin (ie, PZI or ultralente), the duration of action, as determined from the serial blood glucose curve, will determine whether once or twice daily administration is required. Twice-daily administration is indicated if either: (1) the blood glucose drops to a nadir and returns to within 60–70% of the baseline and more than 12 mmol/l by 12 h after injection, or (2) the blood glucose level drops to a distinct nadir then rises back above 15 mmol/l by 12 h after injection. When twice daily administration is required, lente insulin may be preferred, as it has been found in human diabetics that twice daily ultralente leads to a higher incidence of hypoglycaemic episodes than twice daily lente (Tunbridge et al 1989).

If lente or NPH insulin is being used and diabetic control is poor, two circumstances may warrant a change to longer acting insulin (ultralente). Firstly, if the insulin is having a rapid peak of action (within 2–3 h) and a short duration of effect (6–7 h), and/or secondly if a low nadir blood glucose of (<3.5 mmol/l) is occurring despite the mean blood glucose remaining high (above 15 mmol/l).

Once clinical signs are well controlled, and the insulin dose is unchanged on two consecutive visits, then re-evaluation can be performed at longer intervals (every 3–4 months) unless a change in clinical status or hypoglycaemia occurs.

Monitoring diabetics . In many cats, good clinical control can be attained even if there is still hyperglycaemia (eg, mean blood glucose ≥12 mmol/l). Evidence for good clinical control includes an active cat with a healthy appearance, a stable, normal bodyweight, and levels of polydipsia and polyuria acceptable to the owner and veterinarian. Achieving normoglycaemia for most of the day is unlikely in all but a small proportion of diabetic cats (Goossens et al 1998). Water intake can be measured at home or in hospital, and correlates moderately well with the mean blood glucose (Martin & Rand 1999a), providing a useful clinical indicator of diabetic status. In particular, changes in the water intake in an individual cat over time can give clues as to the success (or failure) of treatment. However, water intake may vary substantially from day to day, so measurement for more than a single day, and preferably over a week, should be performed if insulin adjustments are to be based on water intake. A water intake of less than 20 ml/kg bodyweight/24 h indicates exemplary diabetic control. Most cats with good clinical control drink less than 80 ml/kg bodyweight/24 h. Other clinical signs such as a stable normal bodyweight and the presence or absence of lethargy should also be monitored.

The glycated proteins fructosamine and glycated haemoglobin (HbA_1c) have been advocated as indicators of diabetic control in cats (Crenshaw et al 1996, Thoresen & Bredal 1996, Elliott et al 1997). When fructosamine is being used as an indicator of glycaemic control, it should be measured when the cat is normally hydrated to avoid inaccurate results (Link & Rand, unpublished data). Although the means of glycated haemoglobin and fructosamine show significant differences between cats which were grouped into normals, well controlled diabetics, poorly controlled diabetics, and untreated diabetics, the results for individual cats may overlap between the groups (Crenshaw et al 1996, Elliott et al 1997). This makes the interpretation of glycaemic control for an individual cat based on fructosamine alone difficult. Fructosamine levels may be normal in cats with hyperglycaemia, as it was found that normal cats infused with glucose to create hyperglycaemia of around 20 mmol/l did not have fructosamine levels above the reference range (Link & Rand 1995).

High fructosamine levels (>500 μmol/l) indicate a problem, but not necessarily whether the insulin dose needs to be increased or decreased. Making insulin dose rate changes on the basis of fructosamine measurements should be done with caution, as the fructosamine level does not give an indication of the nadir blood glucose. However, fructosamine may be a useful marker where stress or fractiousness makes an accurate serial blood glucose curve unobtainable. As with water intake, fructosamine is probably most useful for monitoring change in an individual diabetic cat over time. There are no studies that demonstrate convincingly that glycated proteins are better, or even as useful as clinical parameters and blood glucose measurement for monitoring diabetic control in cats.

Urine glucose measurements are an inexpensive and simple method for monitoring of diabetic status by the owner, and can be measured from urine soaked cat litter if necessary (Schaer 1994). However, urine glucose also does not give information about how an insulin dose should be changed, and is most useful for indicating or predicting diabetic remission. We get owners to measure urine glucose once or twice weekly if possible, particularly in the first 3–6 months of therapy. If the cat becomes aglycosuric, this may indicate that diabetic remission has occurred, and the insulin can be discontinued for several days while the urine glucose is monitored daily. If glycosuria recurs, it may be necessary to perform a serial blood glucose curve to determine the subsequent insulin dose. Substantial worsening of glycosuria in a diabetic cat indicates that the cat should be re-evaluated using blood glucose measurements to determine the direction of adjustment of insulin dose.

Owner initiated dose rate changes and home monitoring . As the blood glucose response to insulin is so variable, and individual blood glucose measurements may be easily influenced by stress, owner monitoring of blood glucose may give a better indication of true glycaemic control, especially if curves are performed frequently, for example weekly. However, owner-initiated insulin dose rate changes are not recommended and insulin dose rate adjustments should follow discussion with the veterinarian managing the case. We advise owners to monitor urine glucose, bodyweight, and the cat's food and water intake as a measure of diabetic control. If significant changes occur, or if markedly abnormal values persist, a serial blood glucose curve should be performed before insulin dose rate changes are made. Cats may have marked polydipsia and glycosuria while having a low blood glucose nadir, so increasing the insulin dose without performing serial blood glucose measurements risks a hypoglycaemic episode. If blood glucose is only measured in the afternoon when using twice daily insulin, especially lente or NPH insulin, the nadir glucose may be missed. This may lead to inappropriate dose increases and a risk of hypoglycaemia. If only a few blood glucose measurements are performed for cost reasons, the best information is gained from glucose measurements around the time of, and including the glucose nadir. With lente insulin, the nadir occurs approximately 4 h after insulin administration. With PZI insulin, the time of nadir glucose is very variable. For lente insulin, glucose measurements at 2, 4, and 6 h may be sufficient to make an informed adjustment of insulin dose, but measurements every 2 h from 0 to 8 or 12 h after insulin administration gives better information. For PZI, glucose should be measured for at least 6, 8, 12, and 14 h after administration. However, better information is gained by including 0, 2 and 4 h measurements.

Concomitant disease

Several diseases which cause insulin resistance may underlie or complicate the treatment of diabetes; in particular hyperthyroidism, hyperadrenocorticism, acromegaly, infections, pancreatitis, and uraemia from renal failure (Ihle & Nelson 1991, Peterson 1995). Bacterial cystitis, which occurs with a high incidence in diabetic cats especially if polyuria is present (Crenshaw & Peterson 1996, Kirsch 1998), should be treated with appropriate antibiotics. Although there have been no detailed studies, anecdotal reports indicate that diabetic cats commonly have dental disease and chronic gingivitis (Diehl 1995). The authors have encountered insulin-treated diabetic cats in which diabetic control improved markedly when the dental disease was treated. It has been recorded in other species that chronic inflammatory disease may lead to a chronic hypercortisolaemia (Ley et al 1994), which may reduce insulin sensitivity and therefore in diabetics decrease insulin responsiveness. Although treatment of concurrent disease will not improve glycaemic control in some cats (Goossens et al 1998), it is recommended that chronic disorders of this type are addressed as part of the diabetes treatment regime.

Survival

In one study of 92 diabetic cats, average survival time was 17 months (Goossens et al 1998). The survival of diabetic cats is affected by old age and concurrent disease, as the diabetic cats in this study had a mean age of 12 years (Goossens et al 1998). A significant proportion of cats diagnosed with diabetes survive less than 12 months (Kraus et al 1997, Goossens et al 1998). Factors reducing survival time included diabetic ketoacidosis at diagnosis, poor glycaemic control, and the presence of concurrent disease (principally renal failure) (Kraus et al 1997, Goossens et al 1998).

Concluding remarks

Through an understanding of the pathophysiology of feline diabetes, cats can be treated more easily and successfully. Studies demonstrating the lack of post-prandial hypoglycaemia in cats have highlighted an important species difference, that shows ad libitum feeding is suitable for diabetic cats. Glipizide remains the only oral agent that has been studied in detail in diabetic cats. Preliminary studies have been performed on other agents, but further clinical studies in diabetic cats are necessary to determine the efficiency of these drugs. Currently, insulin provides the most successful clinical and glycaemic control for the majority of diabetic cats. Successful treatment requires an understanding of the advantages and limitations of different treatment modalities, and the rationale for clinical decision making and therapeutic monitoring.

Summary: Blood glucose curves and changing the insulin dose

Measuring

Ensure that stress and struggling are minimised.

Use venous catheters or vacuum assisted lances to obtain multiple samples atraumatically.

Pocket blood glucose meters are suitable for performing blood glucose curves in a clinic or home setting.

Feed as usual.

Measure blood glucose immediately before insulin is given, then every 2 h.

Preferably follow the blood glucose through to the next insulin dose (12 or 24 h later).

If this is not possible, follow the blood glucose until it comes back close to baseline.

If the blood glucose drops below 3.0 mmol/l, take more frequent samples to ensure hypoglycaemia is avoided. Administer oral or parenteral glucose as needed.

Interpreting

Parameters that can be determined from the serial blood glucose curve include baseline blood glucose concentration, blood glucose concentration at nadir, time to reach nadir of blood glucose, and the mean blood glucose. The difference between baseline and nadir may be measured, as well as the time taken for the blood glucose to return to baseline concentration.

The nadir of blood glucose concentration is the most important parameter for determining insulin dose rate changes, as it determines whether the dose can be safely increased, or if the dose needs to be reduced.

The time taken to reach blood glucose nadir and the time taken to return to baseline blood glucose concentration are used to determine the dose frequency and the insulin type.

The difference between baseline and nadir of blood glucose concentration can be used to determine how responsive the cat is to given doses of insulin.

The mean blood glucose gives a measure of the overall diabetic control.

Insulin and dose changes

Twice-daily administration is indicated if either: (1) the blood glucose drops to a nadir and returns to within 60–70% of the baseline and more than 12 mmol/l by 12 h after injection, or (2) the blood glucose level drops to a distinct nadir then rises back above 15 mmol/l by 12 h after injection

If the nadir of blood glucose is occurring within 2 or 3 h of injection, a longer acting insulin should be considered.

Insulin dose rate changes should be made conservatively. Where indicated, dose increases should be made at 2–4 week intervals and 1 unit at a time.

The target for treatment is to achieve a blood glucose concentration of 5.0–9.0 mmol/l at nadir. If the nadir blood glucose is greater than 9.0 mmol/l, the insulin dose should be increased by 1 iu.

Daily changes in insulin dose based on food intake are not recommended. If the cat shows a reduced or lack of appetite for more than 12–24 h at home, the veterinarian should examine the cat.

References

Anderson

Akanji

(1991) Dietary fiber—an overview. Diabetes Care 14, 1126–1131.

Appleton

Rand

Priest

Sunvold

(1999) Reference values for glucose tolerance, insulin tolerance, and insulin sensitivity tests in normal cats (Abstract). Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, p. 375.

Arnold

(1996) Non-invasive glucose monitoring. Current Opinion in Biotechnology 7, 46–49.

Bertoy

Nelson

Feldman

(1995) Effect of lente insulin for treatment of diabetes mellitus in 12 cats. Journal of the American Veterinary Medical Association 206, 1729–1731.

Binder

(1969) Absorption of injected insulin: A clinical-pharmacological study (Thesis). Acta Pharmacologica et Toxicologica 27 (Suppl 2), 1–87.

Binder

Lauritzen

Faber

Pramming

(1984) Insulin pharmacokinetics. Diabetes Care 7, 188–199.

Biourge

Nelson

Feldman

Willits

Morris

Rogers

(1997) Effect of weight gain and subsequent weight loss on glucose tolerance and insulin response in healthy cats. Journal of Veterinary Internal Medicine 11, 86–91.

Boudinot

Ferguson

Hoenig

(1999) Pharmacokinetics of troglitazone, an insulin-senstitising agent, in healthy cats (Abstract). Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, p. 708.

Brange

Owens

Kang

Volund

(1990) Monomeric insulins and their experimental and clinical implications. Diabetes Care 13, 923–954.

10.

Broussard

Peterson

(1994) Comparison of two ultralente insulin preparations with protamine zinc insulin in clinically normal cats. American Journal of Veterinary Research 55, 127–131.

11.

Broussard

Wallace

(1995) Insulin treatment of diabetes mellitus in the dog and cat. In: Kirk's Current Veterinary Therapy Bonagura

(ed). Philadelphia, USA, WB Saunders, pp. 393–398.

12.

Bruskiewicz

Nelson

Feldman

Griffey

(1997) Diabetic ketosis and ketoacidosis in cats: 42 cases (1980–1995). Journal of the American Veterinary Medical Association 211, 188–192.

13.

Burrows

(1973) Techniques and complications of intravenous and intraarterial catheterization in dogs and cats. Journal of the American Veterinary Medical Association 163, 1357–1363.

14.

Casella

Mongilio

Plotnick

Hesterberg

Long

(1993) Accuracy and precision of low-dose insulin administration. Pediatrics 91, 1155–1157.

15.

Chantelau

Sonneberg

Rajab

Romisch

Berger

(1985) Absorption of subcutaneously administered regular human and porcine insulin in different concentrations. Diabete et Metabolisme 11, 106–110.

16.

Chow

C-C

Tsang

LWW

Sorenson

J-P

Cockram

(1995) Comparison of insulin with or without continuation of oral hypoglycemic agents in the treatment of secondary failure in NIDDM patients. Diabetes Care 18, 307–314.

17.

Crenshaw

Peterson

(1996) Pretreatment clinical and laboratory evaluation of cats with diabetes mellitus: 104 cases (1992–1994). Journal of the American Veterinary Medical Association 209, 943–949.

18.

Crenshaw

Peterson

Heeb

Moroff

Nichols

(1996) Serum fructosamine concentration as an index of glycemia in cats with diabetes mellitus and stress hyperglycemia. Journal of Veterinary Internal Medicine 10, 360–364.

19.

Diehl

(1995) Long-term complications of diabetes mellitus, part II: Gastrointestinal and infectious. Veterinary Clinics of North America. Small Animal Practice 25, 731–752.

20.

Dow

Lecouteur

Fettman

Spurgeon

(1987) Potassium depletion in cats. Journal of the American Veterinary Medical Association 191, 1563–1568.

21.

Dowling

(1995) Insulin therapy for dogs and cats. Canadian Veterinary Journal 36, 577–579.

22.

Eaton

Spencer

Schade

Shafer

Corbett

(1978) Diabetic glucose control: Matching plasma insulin concentration to dietary and stress hyperglycemia. Diabetes Care 1, 40–44.

23.

Elliott

Nelson

Feldman

Neal

(1997) Glycosylated hemoglobin concentration for assessment of glycemic control in diabetic cats. Journal of Veterinary Internal Medicine 11, 161–165.

24.

Feldhahn

Rand

Martin

(1998) Insulin sensitivity in normal and diabetic cats, and normal cats under stress (Abstract). 16th Annual Veterinary Medical Forum of the American College of Veterinary Internal Medicine, San Diego, CA, USA.

25.

Feldman

Nelson

Feldman

(1997) Intensive 50-week evaluation of glipizide administration in 50 cats with previously untreated diabetes mellitus. Journal of the American Veterinary Medical Association 210, 772–777.

26.

Feldman

Nelson

, (1996) Canine and Feline Endocrinology and Reproduction (2nd edn). WB. Saunders, Philadelphia.

27.

Fondacaro

Greco

Crans

(1999) Treatment of feline diabetes mellitus with protamine zinc alanine insulin alone compared with PZI and oral vanadium salts (Abstract). Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, p. 710.

28.

Forrester

Moreland

(1989) Hypophosphatemia. Causes and clinical consequences. Journal of Veterinary Internal Medicine 3, 149–159.

29.

Galloway

Spradlin

Nelson

Wentworth

Davidson

Swarner

(1981) Factors influencing the absorption, serum insulin concentration, and blood glucose after injections of regular insulin and various insulin mixtures. Diabetes Care 4, 366–376.

30.

Garvey

Olefsky

Griffin

Hamman

Kolterman

(1985) The effect of insulin treatment on insulin secretion and action in type 2 diabetes mellitus. Diabetes 34, 222–234.

31.

Gerich

(1988) Lilly lecture 1988. Glucose counterregulation and its impact on diabetes mellitus. Diabetes 37, 1608–1617.

32.

Goossens

MMC

Nelson

Feldman

Griffey

(1998) Response to treatment and survival in 104 cats with diabetes mellitus (1985–1995). Journal of Veterinary Internal Medicine 12, 1–6.

33.

Greco

(1997) Treatment of non-insulin dependent diabetes mellitus with oral hypoglycemic agents. 15th Annual Forum of the American College of Veterinary Internal Medicine, Lake Buena Vista, Florida, USA, pp. 252–254.

34.

Greco

(1999) Oral hypoglycemic therapy in cats. Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, pp. 647–649.

35.

Greco

Broussard

Peterson

(1995) Insulin therapy. Veterinary Clinics of North America. Small Animal Practice 25, 677–690.

36.

Hallden

Gafvelin

Mutt

Jornvall

(1986) Characterization of cat insulin. Archives of Biochemistry and Biophysics 247, 20–27.

37.

Hand

Armstrong

Allen

(1989) Obesity: Occurrence, treatment, and prevention. Veterinary Clinics of North America. Small Animal Practice 19, 447–474.

38.

Harb-Hauser

Nelson

Gershwin

Neal

(1998) Prevalence of insulin antibodies in diabetic cats (Abstract). 16th Annual Forum of the American College of Veterinary Internal Medicine, San Diego, USA, p. 735.

39.

Heinemann

Chanteleau

Starke

AAR

(1992) Pharmacokinetics and pharmacodynamics of subcutaneously administered U40 and U100 formulations of regular human insulin. Diabete et Metabolisme 18, 21–24.

40.

Hildebrandt

(1991) Subcutaneous absorption of insulin in insulin-dependent diabetic patients. Danish Medical Bulletin 38, 337–346.

41.

Hoenig

O'Brien

(1998) Glipizide leads to amyloidosis in a cat model of type 2 diabetes (Abstract). Diabetologia (Suppl. 1), A648.

42.

Home

Alberti

KGGM

(1992) Insulin therapy. In: International Textbook of Diabetes Mellitus, Alberti

KGGM

DeFronzo

Keen

Zimmet

(ed.), Vol. 1. John Wiley and Sons, Chichester, NY, pp. 831–863.

43.

Houpt

(1991) Ingestive behaviour: Food and water intake. In: Domestic Animal Behavior for Veterinarians and Animal Scientists, Houpt

(ed.). Iowa State University Press, Ames, Iowa, USA, 270–306.

44.

Ihle

Nelson

(1991) Insulin resistance and diabetes mellitus. Compendium on Continuing Education for the Practicing Veterinarian 13, 197–203.

45.

Kane

(1989) Feeding behaviour of the cat. In: Nutrition of the Dog and Cat. Waltham Symposium 7, Kronfeld

Adkins

Downey

Burger

Rivers

(ed.). Cambridge University Press, Cambridge, UK, pp. 147–158.

46.

Kidson

(1993) The Somogyi effect. Has it ever existed and what harm has it caused? Medical Journal of Australia 159, 480–482.

47.

Kienzle

(1994) Blood sugar levels and renal sugar excretion after the intake of high carbohydrate diets in cats. Journal of Nutrition 124, 2563S–2567S.

48.

Kinnaird

Rand

Baglioni

Jr Blackshaw

(1998) Stress hyperglycaemia in cats (Abstract). 16th Annual Veterinary Medical Forum of the American College of Veterinary Internal Medicine, San Diego, USA, p. 703.

49.

Kirsch

(1998) [Incidence of bacterial cystitis in diabetic dogs and cats at the time of diagnosis. Retrospective study for the period 1990–1996]. Tierarztliche Praxis Ausgabe K (Klientiere und Heimtiere) 26, 32–36.

50.

Kraus

Calvert

Jacobs

Brown

(1997) Feline diabetes mellitus: A retrospective mortality study of 55 cats (1982–1994). Journal of the American Animal Hospital Association 33, 107–111.

51.

Laflamme

Cornelius

Roberts

(1993) Enteral nutrition as an adjunct to the management of a complicated case of diabetes mellitus in a cat. Journal of the American Animal Hospital Association 29, 264–266.

52.

Ley

Waterman

Livingston

Parkinson

(1994) Effect of chronic pain associated with lameness on plasma cortisol concentrations in sheep: A field study. Research in Veterinary Science 57, 332–335.

53.

Link

KRJ

Rand

(1995) Changes in fructosamine concentration in cats with induction and resolution of chronic hyperglycemia (abstract). Journal of Veterinary Internal Medicine 9, 180.

54.

Link

KRJ

Rand

(1996) Glucose toxicity in cats (abstract). Journal of Veterinary Internal Medicine 10, 185.

55.

Link

KRJ

Rand

Hendrikz

(1997) Evaluation of a simplified intravenous glucose tolerance test and a reflectance glucose meter for use in cats. Veterinary Record 140, 253–256.

56.

Macintire

(1995) Emergency therapy of diabetic crises: Insulin overdose, diabetic acidosis, and hyperosmolar coma. Veterinary Clinics of North America. Small Animal Practice 25, 639–650.

57.

Martin

Rand

(1997) Lack of correlation between feeding and blood glucose in diabetic cats (Abstract). 15th Annual Forum of the American College of Veterinary Internal Medicine, Orlando, Florida, USA.

58.

Martin

GJW

Rand

(1999a) Correlations between monitoring parameters for insulin-treated diabetic cats (Abstract). Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, p. 735.

59.

Martin

GJW

Rand

(1999b) Evaluation of a polyurethane jugular catheter in cats placed using a modified Seldinger technique. Australian Veterinary Journal 77, 250–254.

60.

McMillan

Feldman

(1986) Rebound hypoglycemia following overdosing of insulin in cats with diabetes mellitus. Journal of the American Veterinary Medical Association 188, 1426–1431.

61.

Michels

Boudinot

Ferguson

Hoenig

(1999) Pharmacokinetics of the antihyperglycemic drug metformin in cats. Journal of Veterinary Internal Medicine 60, 738–742.

62.

Miller

(1995) Long-term monitoring of the diabetic dog and cat: Clinical signs, serial blood glucose determinations, urine glucose, and glycated blood proteins. Veterinary Clinics of North America. Small Animal Practice 25, 571–584.

63.

Miller

Nelson

Kirk

Neal

Feldman

(1992) Effect of glipizide on serum insulin and glucose concentrations in healthy cats. Research in Veterinary Science 52, 177–181.

64.

Moise

Riemers

(1983) Insulin therapy in cats with diabetes mellitus. Journal of the American Veterinary Medical Association 182, 158–164.

65.

Morris

Rogers

(1983) Nutritional implications of some metabolic anomalies of the cat. Proceedings of the 50th Annual Meeting of the American Animal Hospital Association, USA, pp. 325–331.

66.

Munana

(1995) Long-term complications of diabetes mellitus, part I: Retinopathy, nephropathy, neuropathy. Veterinary Clinics of North America. Small Animal Practice 25, 715–730.

67.

Nelson

Feldman

Ford

Kirk

(1992a) Transient diabetes mellitus in the cat (Abstract). Journal of Veterinary Internal Medicine 6, 111.

68.

Nelson

Scott-Montcrief

DeVries

Davenport

Neal

(1994) Dietary insoluble fiber and glycemic control of diabetic cats (Abstract). Proceedings of the 12th ACVIM Forum, San Francisco, CA, USA, p. 996.

69.

Nelson

Feldman

DeVries

(1992b) Use of ultralente insulin in cats with diabetes mellitus. Journal of the American Veterinary Medical Association 200, 1828–1829.

70.

Nelson

Feldman

Ford

Roemer

(1993) Effect of an orally administered sulphonylurea, glipizide, for treatment of diabetes mellitus in cats. Journal of the American Veterinary Medical Association 203, 821–827.

71.

Nelson

Himsel

Feldman

Bottoms

(1990) Glucose tolerance and insulin response in normal weight and obese cats. American Journal of Veterinary Research 51, 1357–1362.

72.

Neubauer

Schone

(1978) The immunogenicity of different insulins in several animal species. Diabetes 27, 8–15.

73.

Nicholls

Crenshaw

(1995a) Complications and concurrent disease associated with diabetic ketoacidosis and other severe forms of diabetes mellitus. In: Kirk's Current Veterinary Therapy, Bonagura

(ed.). Philadelphia, USA, WB Saunders, pp. 384–387.

74.

Nicholls

Crenshaw

(1995b) Complications and concurrent disease associated with diabetic ketoacidosis and other severe forms of diabetes mellitus. Veterinary Clinics of North America. Small Animal Practice 25, 617–624.

75.

Nijs

HGT

Radder

Frolich

Krans

HMJ

(1989) The course and determinants of insulin action in type 1 (insulin-dependent) diabetes mellitus. Diabetologia 32, 20–27.

76.

Peterson

(1995) Diagnosis and management of insulin resistance in dogs and cats with diabetes mellitus. Veterinary Clinics of North America. Small Animal Practice 25, 691–714.

77.

Rachman

Payne

Levy

Barrow

Holman

Turner

(1998) Changes in amylin and amylin-like peptide concentrations and beta-cell function in response to sulfonylurea or insulin therapy in NIDDM. Diabetes Care 21, 810–816.

78.

Rand

(1997) Management of feline diabetes. Australian Veterinary Practitioner 27, 68–78.

79.

Rand

(1999) Current understanding of feline diabetes: Part 1, pathogenesis. Journal of feline Medicine and Surgery 1, 143–153.

80.

Reichard

(1994) A risk-benefit assessment of conventional versus intensive insulin therapy. Drug Safety 10, 196–202.

81.

Schade

Burge

(1995) Brittle diabetes: Etiology and treatment. Advances in Endocrinology and Metabolism 6, 289–319.

82.

Schaer

(1983) Insulin treatment for the diabetic dog and cat. Compendium on Continuing Education for the Practicing Veterinarian 5, 579–588.

83.

Schaer

(1994) A method for detecting glycosuria in urine soaked cat litter. Feline Practice 22, 6–9.

84.

Schernthaner

(1985) Improvement of insulin action is an important part of the antidiabetic effect of metformin. Hormone and Metabolic Research. Supplement 15, 116–120.

85.

Schernthaner

(1993) Immunogenicity and allergic potential of animal and human insulins. Diabetes Care 16 (Suppl 3), 155–165.

86.

Skyler

(1988) Insulin pharmacology. Medical Clinics of North America 72, 1337–1354.

87.

Sloth

(1992) The practical management of obesity in cats and dogs. Journal of Small Animal Practice 33, 178–182.

88.

Steil

Murray

Bergman

Buchanan

(1994) Repeatability of insulin sensitivity and glucose effectiveness from the minimal model. Implications for study design. Diabetes 43, 1365–1371.

89.

Thoresen

Bredal

(1996) Clinical usefulness of fructosamine measurements in diagnosing and monitoring feline diabetes mellitus. Journal of Small Animal Practice 37, 64–88.

90.

Tunbridge

FKE

Newens

Home

Davis

Murphy

Burrin

Alberti

KGGM

Jensen

(1989) A comparison of human ultralente- and lente-based twice-daily injection regimens. Diabetic Medicine 6, 496–501.

91.

Wess

Reusch

(1999) New methods for simplified glucose measurements by capillary blood sampling (Abstract). Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, p. 734.

92.

Whitley

Drobatz

Panciera

(1997) Insulin overdose in dogs and cats: 28 cases (1986–1993). Journal of the American Veterinary Medical Association 211, 326–330.

93.

Willard

Zerbe

Schall

Johnson

Crow

Jones

(1987) Severe hypophosphatemia associated with diabetes mellitus in six dogs and one cat. Journal of the American Veterinary Medical Association 190, 1007–1010.

94.

Zerbe

(1999) The glucose curve: what is it and how should we use it? Proceedings of the 17th ACVIM Forum, Chicago, IL, USA, pp. 436–437.

95.

Ziel

Davidson

Harris

Rosenberg

(1988) The variability in the action of unmodified insulin is more dependent on changes in tissue sensitivity than on insulin absorption. Diabetic Medicine 5, 662–666.

96.

Zimmerman

(1997) Sulfonylureas. Endocrinology and Metabolism Clinics of North America 26, 511–522.

Current Understanding of Feline Diabetes: Part 2,Treatment

Abstract

Sick ketotic diabetic cats

Therapeutic goals for otherwise healthy diabetic cats

Treatment modalities

Diet

Oral hypoglycaemics

Insulin

Concomitant disease

Survival

Concluding remarks

Summary: Blood glucose curves and changing the insulin dose

Measuring

Interpreting

Insulin and dose changes

References