Diabetes in Dogs

Diabetes mellitus is a common endocrine disease of dogs and requires life-long therapy. Nutritional management is an important part of the treatment regimen and feeding guidelines based on evidence from well-designed clinical studies are essential. The first part of this chapter provides an understanding of the pathogenesis of diabetes in dogs, which is required before evaluation of issues relating to nutritional management. This allows comparison of the current, evidence-based, nutritional recommendations for human patients with types of diabetes analogous to canine diabetes. The second part reviews the available evidence from feeding studies in dogs and provides detailed analysis of the recommendations for dietary fiber, carbohydrate, fat, protein, and selected micronutrients in diabetic dogs. The final summary uses the American Diabetes Association grading system to rank the scientific basis of the nutritional recommendations for canine diabetes.

Linda FLEEMAN
BVSc, MACVSc

Linda Fleeman graduated with honors from the University of Queensland and completed Clinical Residency training programs in Small Animal Medicine at both Murdoch University and the University of Melbourne in Australia. She is currently participating in a clinical PhD research project investigating the therapy and nutritional management of diabetes in dogs. She is currently Lecturer in Small Animal Medicine at the University of Queensland. Dr Fleeman is first author of a number of publications focusing on canine diabetes, and has lectured at many national and international conferences on this topic.

Jacquie RAND
BVSc (Hons), DVSc, Dipl ACVIM

Professor Rand graduated from the University of Melbourne (Australia) in 1975 and worked in private practice for 8 years before completing a residency and doctorate at the University of Guelph (Canada). She is currently Professor of Companion Animal Health at the University of Queensland and Director of the Centre for Companion Animal Health. Jacquie is recognized internationally as a leader in feline diabetes and nutrition research. She has authored over 100 articles, seventy abstracts and six textbook chapters. She currently has a team of 10 postgraduate students working in diabetes, obesity and nutrition research in companion animals.

1. Diabetes in Dogs

Prevalence of Diabetes Mellitus in Dogs

Diabetes mellitus is one of the most frequent endocrine diseases affecting middle-aged and older dogs, and the prevalence is increasing. Thirty years ago, 19 in 10,000 dogs visiting veterinary hospitals were diagnosed with diabetes mellitus (Marmor et al., 1982; Guptill et al., 2003). By 1999, the prevalence in the same veterinary hospitals had increased three-fold to 58 per 10,000 dogs (Figure 1) (Guptill et al., 2003).

Figure 1. Increasing prevalence of canine diabetes mellitus. Reprinted from Guptill et al., 2003, with permission from Elsevier.

Clinical Sequelae of Diabetes in Dogs

Insulin deficiency results in altered carbohydrate, fat, and protein metabolism. Abnormal carbohydrate metabolism manifests as hyperglycemia and glycosuria and is responsible for the polyuria, polydipsia, and cataract formation seen in diabetic dogs. The hyperlipidemia, ketone production, and hepatic changes seen in these dogs primarily results from altered fat metabolism. Decreased tissue utilization of glucose, amino acids, and fatty acids causes lethargy, weight loss, reduced stimulation of the satiety center, poor coat, and reduced immunity that is characteristic of untreated diabetic dogs.

Cataract Formation

Cataract formation is the most common, and one of the most important, long-term complications associated with diabetes in dogs (Beam et al., 1999) (Figure 2a and Figure 2b). Cataracts are irreversible and can progress quite rapidly (Figure 3a-Figure 3c). About 30% of diabetic dogs already have reduced vision at presentation (Graham & Nash, 1997a). Cataracts will develop within 5 - 6 months of diagnosis in the majority of diabetic dogs and, by 16 months, approximately 80% will have significant cataract formation (Beam et al., 1999). Importantly, the risk of cataract development seems to be unrelated to the level of hyperglycemia but increases with age (Salgado et al., 2000). Thus, dietary manipulation is not likely to influence the rate or severity of cataract development in diabetic dogs.

Figure 2a. Diabetic cataract associated with uveitis in a dog. Advanced cataract in an aging dog. Hyperemia is present in the sclera, indicating moderate uveitis. (©RIE Smith).

Figure 2b. Diabetic cataract associated with uveitis in a dog. Severe uveitis in a diabetic dog. The eye is red and painful, with the presence of mucopurulent ocular discharge and posterior synechia.

Figure 3a. Development of diabetic cataracts in a dog (from Fleeman & Rand 2000). An eleven-year-old crossbred dog photographed shortly after diagnosis of diabetes mellitus. (©RIE Smith).

Figure 3b. Development of diabetic cataracts in a dog. The same dog three months after initial diagnosis of diabetes mellitus. Diabetic cataracts have rapidly developed and the dog’s owner reported sudden vision loss. (©RIE Smith).

Figure 3c. Development of diabetic cataracts in a dog. The same dog following phacoemulsification surgery to remove the cataract from the right eye. (©RIE Smith).

Concurrent Disease

Treated diabetic dogs have a similar chance of survival as compared to non-diabetic dogs of the same age and gender, although the hazard of death occurring is greatest during the first 6 months of therapy (Graham & Nash, 1997b). Most diabetic dogs are middle-aged and older and are prone to diseases that commonly affect this age group. Consequently, many suffer concurrent problems that need to be managed in combination with the diabetes. For diabetic dogs receiving insulin therapy, the nutritional requirements of any concurrent disease may need to take precedence over the dietary therapy for diabetes. Regardless of the diet fed, glycemic control can still usually be maintained with exogenous insulin therapy.

If concurrent illness causes transient inappetence, it is generally advisable to administer half the usual insulin dose to reduce the risk of hypoglycemia. Diabetic dogs with a reduced appetite will often eat if they are hand-fed highly palatable food by their owner. If there is a more severe concurrent illness causing prolonged inappetence, diabetic dogs should be hospitalized for blood glucose concentration monitoring and treatment with a rapid-acting insulin preparation and intravenous fluids supplemented with glucose and potassium (Feldman et al., 2004a).

Insulin-induced Hypoglycemia

Severe hypoglycemia resulting from insulin overdose can cause irreversible brain damage and death, and avoidance of insulin-induced hypoglycemia is one of the primary aims of therapy in diabetic dogs. Dietary manipulation that reduces the risk of insulin-induced hypoglycemia affords important clinical benefit for diabetic dogs. Severe hypoglycemia has been reported in a diabetic dog that was fed ad libitum and received insulin at grossly irregular intervals (Whitley et al., 1997). Commercial dog foods usually result in postprandial elevation of plasma glucose for less than 90 minutes following consumption by dogs (Nguyen et al., 1998a) and meals should ideally be timed so that maximal exogenous insulin activity occurs during the postprandial period (Church, 1982). Thus, dogs should be fed within 2 hours of administration of lente insulin or within 6 hours of protamine zinc insulin (Stenner et al., 2004) (Figure 4). In practice, a feasible compromise is to feed the dog immediately following the insulin injection. This considerably simplifies the home treatment regimen for most dog owners while still allowing good glycemic control to be readily achieved. In addition, many owners prefer this regimen because they feel their pet is rewarded for submitting to the injection.

Figure 4. Pharmacodynamics and pharmacokinetics in 9 healthy, non-diabetic dogs following subcutaneous injection of lente (Caninsulin®, Intervet) and protamine zinc insulin (PZI Vet®, Idexx) preparations (from Stenner et al., 2004).

Because the daily insulin-dosing regimen tends to be fixed for diabetic dogs, it is important that a predictable glycemic response is achieved following each meal. Ideally, every meal should contain the same ingredients and calorie content, and should be fed at the same time each day. It is crucial that the diet fed is palatable so that food intake is predictable. The major determinant of the postprandial glycemic response in dogs is the starch content of the meal (Nguyen et al., 1998b) so particular care should be applied to ensure consistent source and content of dietary starch.

The importance of avoiding an insulin overdose cannot be over-emphasized. Every person in the diabetic dog's household needs to be aware of this life-threatening complication, which can rapidly develop into a serious emergency. If some insulin is spilt during the injection it should never be "topped up", even if it appears that the dog has received no insulin. If the owner is ever uncertain about whether or not to give an insulin dose, the safest option is to withhold the injection, as the consequences of missing a single insulin dose are negligible. If mild signs of hypoglycemia develop, the owner should feed a meal of the dog's usual food. If the dog is unwilling or unable to eat, syrup containing a high glucose concentration can be administered orally. Suitable syrups are marketed for use by human diabetics. When the dog recovers, food should be fed as soon as possible. No more insulin should be given to the dog and the owner should discuss the case with a veterinarian before the next injection is due. A 50% reduction in insulin dose is usually recommended in these circumstances.

Successful management of 94% of diabetic dogs is achieved with twice-daily insulin dosing (Hess & Ward, 2000). High doses of insulin and episodes of hypoglycemia are more common in diabetic dogs that receive insulin only once-daily (Hess & Ward, 2000). Although treatment regimens comprising once-daily insulin injections are considered by some to be simpler and more convenient, most of these regimens involve feeding two meals each day, one soon after the insulin injection and another at the time of peak insulin activity about 8 hours later. Given the length of the usual working day, it may actually be more convenient for people to feed the second meal 12 hours after the first. Experienced owners rarely report any difficulty with the administration of insulin injections and, if they are required to be at home to feed the dog, it is little more effort to give the dog an insulin injection at the same time. As a result, many clinicians favor treatment regimens that involve administration of the same dose of insulin along with feeding of the same sized meal every 12 hours.

The owners of diabetic dogs should be aware that a consistent insulin-dosing and feeding routine is optimal. Conservative, fixed-dose, twice-daily insulin therapy, in conjunction with a palatable diet containing a consistent content and source of dietary starch, which is fed at defined times in relation to insulin administration, is likely to be associated with reduced risk of hypoglycemia in diabetic dogs.

Pathogenesis of Diabetes in Dogs

The current classification of human diabetes mellitus is based on pathogenesis, and thus provides a rational foundation for understanding treatment issues. Adoption of these criteria for canine diabetes will afford a similar benefit for veterinarians. Human diabetes is divided into type 1, type 2, other specific types of diabetes, and gestational diabetes (The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, 1997). At present, there are no internationally accepted criteria for the classification of canine diabetes. If the criteria established for human diabetes were applied to dogs, at least 50% of diabetic dogs would be classified as type 1. The remainder are likely to have "other specific types of diabetes" resulting from pancreatic destruction or chronic insulin resistance, or they have diestrus-induced diabetes.

Type 1 Diabetes

Type 1 diabetes appears to be the most common form of diabetes in dogs, and is characterized by pancreatic beta cell destruction leading to absolute insulin deficiency. In people, this usually occurs via cell-mediated, autoimmune processes and is associated with multiple genetic predispositions and poorly defined environmental factors (The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, 1997). The majority of diabetic dogs have absolute insulin deficiency (Montgomery et al., 1996). The etiology of beta cell destruction is often unknown, although there is evidence that in approximately 50% of diabetic dogs it is caused by immune-mediated processes similar to human type 1 diabetes (Alejandro et al., 1988; Hoenig & Dawe; 1992; Davison et al., 2003a, Davison et al., 2003b).

Evidence is mounting for a genetic basis for canine diabetes. An association with major histocompatibility complex alleles on the dog leukocyte antigen gene strongly suggests that the immune response has a role in the pathogenesis of diabetes mellitus (Kennedy et al., 2003; Davison et al., 2003a; Rand et al., 2004).

Although genetic susceptibility appears to be a prerequisite, multiple environmental factors likely initiate beta cell autoimmunity, which once begun, then proceed by common pathogenic pathways (Kukreja & Maclaren, 1999). Similar to canine diabetes, the incidence rate of type 1 diabetes in people is rising (Onkamo et al., 1999), a trend that has been explained on the basis of increased contacts with adverse environmental factors (Kukreja & Maclaren, 1999). There is a highly significant seasonal incidence of diagnosis of both human type 1 diabetes (Gamble & Taylor, 1969; Fleegler et al., 1979) and canine diabetes (Atkins & MacDonald, 1987), with the incidence peaking in winter, indicating that environmental influences may have a role in disease progression just prior to diagnosis.

The rate of progression to absolute insulin deficiency is quite variable in human patients. It can be rapid in young children and much slower in middle-aged and older people. This latter group has the latent autoimmune diabetes of adults (LADA) form of type 1 diabetes, which is characterized by gradual beta cell destruction over months or years and is not associated with obesity (Zimmet et al., 1994). Distinct autoantibody patterns are recognized in the acute onset and slowly progressive (LADA) forms of human type 1 diabetes (Zimmet et al., 1994; Seissler et al., 1998), indicating a different pathogenesis for the two forms of the disease.

The rate of progression to absolute insulin deficiency has not been studied in dogs, but epidemiological factors closely match those of human patients with the LADA form of type 1 diabetes, who are usually not obese and tend to be middle-aged and older. Most affected dogs are over 7 years of age and the onset of clinical signs is typically insidious, ranging from weeks to months in duration (Ling et al., 1977). This has prompted speculation that there may also be similarities between the pathogenesis of canine diabetes and human LADA.

Other Types of Canine Diabetes

Association Between Diabetes and Pancreatitis in Dogs

Extensive pancreatic damage, which likely results from chronic pancreatitis, is responsible for the development of diabetes in approximately 28% of diabetic dogs (Alejandro et al., 1988) and thus is the most common "other specific type" of diabetes in dogs. Beta cell loss is being investigated in non-diabetic dogs with chronic pancreatitis and preliminary findings indicate that some have reduced beta cell function and appear to be pre-diabetic (Watson & Herrtage, 2004). Serum canine pancreatic lipase immunoreactivity (cPLI) is a sensitive marker for pancreatic inflammation in dogs (Steiner, 2003). Increases in serum cPLI concentration have been reported in 5 of 30 (17%) newly diagnosed diabetic dogs, although none of these dogs had serum cPLI concentrations above the diagnostic cut-off value for pancreatitis (Davison et al., 2003b).

In long-term diabetic dogs with no clinical evidence of exocrine pancreatic disease, serum cPLI concentrations in the diagnostic range for pancreatitis were found in 2 of 12 (17%) dogs, with a further 4 (33%) dogs recording increases in cPLI that did not reach the diagnostic cut-off value for pancreatitis, and an additional 2 (17%) dogs having laboratory evidence of exocrine pancreatic insufficiency (unpublished data). This indicates that subclinical exocrine pancreatic disease is common in diabetic dogs.

The association between canine diabetes and pancreatitis warrants particular attention because beta cell autoimmunity, pancreatic inflammation, and regulation of gut immunity might be linked in disease pathogenesis. The gut immune system likely plays a central role in the pathogenesis of human type 1 diabetes because accumulating evidence suggests that affected persons have aberrant regulation of gut immunity (Vaarala, 1999, Akerblom et al., 2002). The gut and the pancreas are probably immunologically linked, as well as anatomically linked, and influenced by environmental factors such as intestinal microflora, infections, and dietary factors (Vaarala, 1999).

Hypertriglyceridemia has been proposed as a possible inciting cause of canine pancreatitis (Williams, 1994) and is commonly seen in diabetic dogs (Ling et al, 1977). Obesity affects one-quarter to one-third of dogs presented to veterinary practices (Edney & Smith, 1986), and is also associated with an increased risk of pancreatitis (Hess et al., 1999). Environmental factors such as feeding high-fat diets, lipemia and disturbances in lipid metabolism, have been implicated as potential etiological factors in dogs with obesity-associated pancreatitis (Simpson, 1993), and likely play a role in the development of pancreatitis in diabetic dogs. More detailed discussion on canine pancreatitis and hyperlipidemia can be found in Chapter 5 and Chapter 7 of this encyclopedia.

Adult Dachshund presenting an excess of weight. No epidemiological data examining the relationship between canine diabetes and obesity have been published since 1960 (Krook et al.), and an association between obesity and diabetes in dogs is not currently recognized. (©Clouquer).

Role of Insulin Resistance in Canine Diabetes

Diabetes induced by insulin resistance states are less common "other specific types" of canine diabetes.

Disease conditions such as hyperadrenocorticism (Peterson, 1984) and acromegaly (Selman et al., 1994) result in insulin resistance, and may induce diabetes in dogs. Iatrogenic causes of insulin resistance that might lead to induced diabetes include chronic corticosteroid therapy (Campbell & Latimer, 1984). As most dogs do not develop overt diabetes with chronic corticosteroid therapy or spontaneous hyperadrenocorticism, for overt diabetes to develop it might require underlying reduced beta cell function resulting from immunological processes or chronic pancreatitis.

Although obesity causes insulin resistance in dogs, there are no published data clearly indicating that obesity is a risk factor for canine diabetes.

Obesity is a well-established risk factor for type 2 diabetes in cats and people. In contrast, there are no well-documented studies demonstrating convincingly that type 2 diabetes is a significant disease entity in dogs. In dogs, obesity causes insulin resistance (Rocchini et al., 1999; Villa et al., 1999, Mittelman et al., 2002), which leads to hyperinsulinemia and impaired glucose tolerance (Mattheeuws et al., 1984, Henegar et al., 2001). These effects are particularly pronounced when obesity is induced by feeding a diet high in saturated fat (Truett et al., 1998). Dogs fed a high-fat diet develop insulin resistance that is not compensated for by increased insulin secretion, resulting in more severe glucose intolerance (Kaiyala et al., 1999). Despite the evidence that obesity causes impaired glucose tolerance, it appears that very few dogs develop overt diabetes as a consequence of obesity-induced insulin resistance.

Diestrus- and Gestation-associated Diabetes

Gestational diabetes is another classification of diabetes recognized in human patients. In women, it is defined as any degree of glucose intolerance with onset or first recognition during pregnancy (The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, 1997). If overt diabetes persists after the pregnancy ends, then it is reclassified as type 1, type 2, or another specific type of diabetes. Reduced insulin sensitivity occurs in healthy bitches by day 30 - 35 of gestation (McCann, 1983) and becomes more severe during late pregnancy (Concannon, 1986). The luteal phase of the non-pregnant cycle of the bitch is similar in duration to the 9 weeks of pregnancy and it is generally agreed that the hormone profiles during diestrus and pregnancy, are essentially identical (Concannon et al., 1989; Feldman et al., 2004b). Progesterone elevation causes glucose intolerance and overt diabetes during diestrus in bitches (Eigenmann et al., 1983, Scaramal et al., 1997). Progesterone also stimulates the mammary gland of bitches to produce growth hormone, which is a potent inducer of insulin resistance (Selman et al., 1994).

The periodic influence of diestrus-associated insulin resistance might contribute to the increased risk of female compared with male dogs for developing diabetes (Marmor et al., 1982; Guptill et al., 2003).

Classification of canine diabetes based on the current understanding pathogenesis is summarized in Table 1.

If diabetes is diagnosed in a bitch during either pregnancy or diestrus, it probably should be classified as of being comparable to human gestational diabetes. If diabetes persists after pregnancy or diestrus ends, then it should be reclassified as type 1 or another specific type of diabetes. (©Lanceau).

Nutritional Perspectives Based on Pathogenesis of Diabetes in Dogs

Understanding the pathogenesis of diabetes in dogs provides a logical foundation for understanding issues relating to nutritional management of this disease. Recently, the American Diabetes Association released a position statement comprising a large meta-analysis presenting evidence-based nutrition principles and recommendations for the treatment and prevention of human diabetes (Franz et al., 2002a). Consideration of the evidence-based recommendations for human patients with types of diabetes comparable to canine diabetes provides a rational perspective for dietary recommendations for diabetic dogs.

Dietary Carbohydrate and Type 1 Diabetes

Perspective gained from the dietary carbohydrate recommendations for human type 1 diabetics provide relevant perspective for canine diabetics because at least 50% of diabetic dogs appear to have analogous disease. Most relevant is perhaps the current recommendation regarding consumption of dietary fiber by human type 1 diabetics. After decades spent researching the effects of dietary fiber on the glycemic and lipemic responses of diabetic people, it is interesting that the current recommendation is that consumption of fiber is to be encouraged in all people and that those with type 1 diabetes require no more dietary fiber than non-diabetic people (Franz et al., 2002a). This suggests that there might also be no clinical benefit of feeding a diet with increased levels of fiber to diabetic dogs compared with feeding "typical", moderate-fiber diets formulated for adult maintenance.

With regard to the glycemic effects of carbohydrates, there is strong evidence in human diabetics that the total amount of carbohydrate in meals and snacks is more important than the source or type (Franz et al., 2002a). Additionally, there is a strong association between the pre-meal insulin dosage required and the postprandial glycemic response to the carbohydrate content of the meal, regardless of the glycemic index, fiber, fat, or caloric content of the meal (Franz et al., 2002a). As a regimen of fixed daily insulin dosages is typically used to manage diabetic dogs, it is rational to provide a consistent amount of carbohydrate in the meals fed each day.

Dietary Fat and Type 1 Diabetes

The primary goal regarding dietary fat in human patients with diabetes is to decrease intake of saturated fat and cholesterol to reduce the risk of coronary heart disease (Franz et al., 2002a). As coronary heart disease is not recognized as a significant clinical entity in dogs, it might not be relevant to extrapolate dietary fat recommendations for human patients to diabetic dogs. For most human type 1 diabetics, effective insulin therapy returns serum lipid levels to normal and usually lowers plasma triglyceride concentrations (Franz et al., 2002a). However, for obese individuals with type 1 diabetes, there is strong evidence that restricted intake of saturated fats, incorporation of monounsaturated fats into the diet, modest weight loss, and increased physical activity may be beneficial (Franz et al., 2002a). The same recommendations might afford clinical benefit for obese diabetic dogs.

Dietary Protein and Type 1 Diabete

The protein composition of the recommended diet for people with diabetes is the same as that recommended for the non-diabetic population (Franz et al., 2002a). However, if microalbuminuria or persistent proteinuria develop, then protein restriction might help slow the progression of diabetic nephropathy in these people (EASD, 1995).

Diabetes with Exocrine Pancreatic Disease

Approximately 60% of human type 1 diabetics have reduced exocrine pancreatic function and it is now recognized that diabetes secondary to exocrine pancreatic disease might be more frequent in people than previously realized (Hardt et al., 2000). Despite this, no specific dietary recommendations are given in the current American Diabetes Association position statement regarding diabetic patients with concurrent exocrine pancreatic disease. Human diabetics with hypertriglyceridemia have increased risk of acute pancreatitis and current management recommendations include a fat-restricted diet (Athyros et al., 2002).

Dietary Recommendations for Gestational Diabetes

In the supplemental American Diabetes Association position statement focusing on gestational diabetes (Franz et al., 2002b), it is noted that restriction of dietary carbohydrate has been shown to decrease maternal postprandial glucose levels (Major et al., 1998). Similarly, bitches with diestrus-associated insulin resistance might benefit from a carbohydrate-restricted diet. This would likely reduce postprandial blood glucose fluctuations, helping to alleviate the hyperinsulinemia associated with diestrus, thus preserving beta cell function and reducing the risk of overt diabetes. There is some evidence that reduced intake of total fat, particularly saturated fat, in people might improve insulin sensitivity and reduce the risk for insulin resistance-associated diabetes (Franz et al., 2002a). Potentially, feeding a fat-restricted diet to bitches with diestrus-associated insulin resistance might improve insulin sensitivity and reduce the risk of overt diabetes. As both fat and carbohydrate restriction may be recommended for these animals, a high-protein diet is a rational choice.

Importantly, nutrient-restricted diets should never be recommended for pregnant bitches unless there is strong scientific evidence for both maternal and fetal benefit.

Dietary Recommendations for Older Diabetics

There are no evidence-based nutritional recommendations for aging diabetic persons and they must be extrapolated from what is known for the general population (Franz et al., 2002a). There is strong evidence that energy requirements for older adults are less than those for younger adults, however it is pointed out that under-nutrition is more likely than over-nutrition in elderly people. Therefore, caution should be exercised when prescribing weight-loss diets (Franz et al., 2002a).

There are no evidence-based nutritional recommendations for aging diabetic dogs. Caution should be exercised when prescribing low-calorie diets to older dogs because this might result in excessive loss of body condition. (©Lanceau).