Canine Diabetic Ketoacidosis
ACVIM 2008
Rebecka S. Hess, DVM, DACVIM
Philadelphia, PA, USA


Diabetic ketoacidosis (DKA) is a severe form of complicated diabetes mellitus (DM) which requires emergency care. Ketones are synthesized from fatty acids as a substitute form of energy, because glucose is not effectively entered into the cells. Excess keto-acids results in acidosis and severe electrolyte abnormalities, which can be life threatening.


Ketone bodies are synthesized as an alternative source of energy, when intracellular glucose concentration can not meet metabolic demands. Ketone bodies are synthesized from acetyl-CoA which is a product of mitochondrial ß-oxidation of fatty acids. Synthesis of acetyl-CoA is facilitated by decreased insulin concentration and increased glucagon concentration. In non-diabetics acetyl-CoA and pyruvate enter the citric acid cycle to form ATP. However, in diabetics, production of pyruvate by glycolysis is decreased. The activity of the citric acid cycle is therefore diminished resulting in decreased utilization of Acetyl-CoA. The net effect of increased production and decreased utilization of acetyl-CoA is an increase in the concentration of acetyl-CoA which is the precursor for ketone body synthesis.1

The three ketone bodies synthesized from acetyl-CoA include beta hydroxybutyrate, acetoacetate, and acetone. Acetoacetate and beta-hydroxybutyrate are anions of moderately strong acids. Therefore, accumulation of these ketone bodies results in ketotic acidosis. Metabolic acidosis and the electrolyte abnormalities which ensue are important determinants in the outcome of patients with DKA.2

One of the beliefs regarding the pathophysiology of DKA had been that individuals that develop DKA have zero or undetectable endogenous insulin concentration. However, in a study that included 7 dogs with DKA it was shown that 5 of 7 dogs with DKA had detectable endogenous serum insulin concentrations, and two of these dogs had endogenous serum insulin concentration within the normal range.3 Therefore, it is possible that other factors, such as elevated glucagon concentration contribute to development of DKA. Glucagon concentration may be elevated due to concurrent disease.

Risk Factors

The median age of dogs with DKA is 8 years (range, 8 months to 16 years).2 The mean age of cats with DKA is 9 years (range, 2 to 16 years).4 Specific breed or sex has not been shown to increase the risk of DKA in dogs or cats.2,4,5

Concurrent disease has been documented in about 70% of dogs with DKA, and in about 90% of cats with DKA. Concurrent disease is a common finding in human beings with DKA. The most common concurrent diseases noted in dogs with DKA are acute pancreatitis, bacterial urinary tract infection, and hyperadrenocorticism.2 The most common concurrent diseases noted in cats with DKA are hepatic lipidosis, chronic renal failure, acute pancreatitis, bacterial or viral infections, and neoplasia.4 It is possible that the presence of concurrent disease results in elevated glucagon concentration and increased risk of DKA.

Most dogs and cats with DKA are newly diagnosed diabetics. It is possible that, as in humans, insulin treatment reduces the risk of DKA in dogs and cats.2,4

Clinical Signs and Physical Examination Findings

Clinical signs and physical examination findings may be attributed to chronic untreated diabetes mellitus, presence of concurrent disease, and the acute onset of DKA. The most common clinical signs of dogs or cats with DKA are polyuria and polydipsia, lethargy, inappetence or anorexia, vomiting, and weight loss.2,4 Common abnormalities noted on physical examination of dogs with DKA are subjectively overweight or underweight body condition, dehydration, cranial organomegaly, abdominal pain, cardiac murmur, mental dullness, dermatologic abnormalities, dyspnea, coughing, or abnormal lung sounds, and cataracts.2 Common abnormalities noted on physical examination of cats with DKA are subjectively underweight body condition, dehydration, icterus, or hepatomegaly.4

Clinical Pathology

Approximately 50% of dogs with DKA have a non-regenerative anemia (which is not associated with hypophosphatemia), left shift neutrophilia, or thrombocytosis.2 Anemia and left shift neutrophilia are also a common feature of feline DKA.4 Cats with DKA also have significantly more red blood cell Heinz body formation compared to normal cats, and the degree of Heinz body formation is correlated with plasma beta-hydroxybutyrate concentration.6

Persistent hyperglycemia is apparent in all dogs and cats diagnosed with DKA, unless they are insulin treated.2 Alkaline phosphatase activity is elevated in almost all dogs with DKA.2 Alanine aminotransferase activity, aspartate aminotransferase activity, or cholesterol concentration are increased in about half of the dogs with DKA.2 Elevations in alanine aminotransferase activity and cholesterol concentration are also commonly observed in cats with DKA.4 Azotemia is reported more commonly in cats with DKA compared to dogs with DKA.2,4

Electrolyte abnormalities are common in both dogs and cats with DKA.2,4 Hypokalemia may develop as excess hydrogen ions are shifted from the extra-cellular fluid into cells. Positively charged potassium ions are then shifted out of cells to compensate for the electric change associated with movement of positively charged hydrogen ions into cells. In addition, these animals often have a decreased intake of potassium due to inappetence or anorexia and increased losses through vomiting and osmotic diuresis. Hyperglycemia and hypoinsulinemia also contribute to a shift of potassium to the extra-cellular fluid. Initially, an animal with DKA may appear to have extra-cellular hyperkalemia due to decreased renal excretion, dehydration, and decreased insulin function. However, with rehydration, potassium ions are lost from the extra-cellular fluid and true hypokalemia becomes apparent. Hypokalemia may be exacerbated by binding of potassium to keto-acids, vomiting, and anorexia. Insulin therapy may worsen extra-cellular hypokalemia as insulin shifts potassium into cells.7 The most important clinical significance of hypokalemia in DKA is profound muscle weakness which may result in respiratory paralysis in extreme cases.

Hypophosphatemia develops when phosphate shifts from the intracellular space to the extracellular space as a result of hyperglycemia, acidosis, and hypoinsulinemia. Osmotic diuresis or fluid therapy along with insulin therapy cause extracellular phosphate depletion leading to whole body phosphate depletion.7 Hypophosphatemia related to DKA has been associated with hemolysis (in a cat) and seizures (in a dog).8 Additional clinical signs that may develop due to hypophosphatemia include weakness, myocardial depression, and arrhythmias.

Decreased plasma ionized magnesium (iMg) concentration has been documented in 4 of 7 cats with DKA, and may be due to increased urinary excretion of magnesium.9 The clinical significance of hypomagnesemia in cats is unknown. The clinical consequence of hypomagnesemia in human diabetics includes insulin resistance, hypertension, hyperlipidemia, and increased platelet aggregation. Dogs with DKA usually do not have low iMg concentration at the time of initial examination.2,10 In one study of 78 dogs with uncomplicated DM, 32 dogs with DKA, and 22 control dogs, plasma iMg2+ concentration at the time of initial examination was significantly higher in dogs with DKA compared to dogs with uncomplicated DM and control dogs.10

Hyponatremia, hypochloremia, and decreased ionized calcium concentration have also been documented in about 50% of dogs with DKA. Low sodium (Na) concentration may be secondary to hyperglycemia, a phenomenon sometimes described as "pseudohyponatremia". In humans, the measurement of hyponatremia can be corrected for hyperglycemia by adding 1.6 mEq/L to the measured sodium for each 100 mg/dl glucose measured above 100 mg/dl, or by adding 1 mEq/L Na for each additional 62 mg/dl of glucose measured above 100 mg/dl.

Urinalysis is usually indicative of glucosuria. Proteinuria or ketonuria may also be apparent. Ketonuria may not be detected because the nitroprusside reagent in the urine dipstick reacts with acetoacetate and not with beta-hydroxybutyrate, which is the dominant ketone body in DKA. Measurement of serum beta-hydroxybutyrate is more sensitive than measurement of urine ketones.11 On urinalysis, the number of white blood cells (WBC) per high power field is usually 5 or less despite the fact that 20% of dogs with DKA have aerobic bacterial growth on urine culture of urine obtained by cystocentesis.2 This is likely due to immunosuppression of diabetics and decreased ability to mobilize WBC to the site of infection.

Results of additional clinicopathologic or imaging tests such as urine culture, abdominal ultrasound, thoracic radiographs, adrenal or thyroid axis testing, pancreatic lipase immunoreactivity, liver function tests, or liver biopsy depend on the presence of specific concurrent disorders.

Differential Diagnosis

Differential diagnosis for ketosis include DKA, acute pancreatitis, starvation, low carbohydrate diet, persistent hypoglycemia, persistent fever, or pregnancy. Differential diagnosis for a primary metabolic acidosis include DKA, renal failure, lactic acidosis, toxin exposure, severe tissue destruction, severe diarrhea, or chronic vomiting.


Administration and careful monitoring of intravenous (IV) fluid therapy is the most important component of treatment. Any commercially available isotonic crystalloid solution may be used. The use of 0.9% saline has been advocated because of its relatively high sodium concentration, and is recommended by the American Diabetes Association for treatment of human beings with DKA.7 However, 0.9% NaCl may be contraindicated in hyperosmolar diabetics. As long as the patient is monitored carefully, in particular in regard to hydration, mental status, and electrolyte concentrations, any crystalloid can be used for treatment. Fluid therapy may contribute to a decrease in blood glucose concentration by improving renal perfusion and decreasing the concentration of counter-regulatory hormones, most importantly, glucagon.12

Correction and monitoring of electrolyte abnormalities is the second most important component of therapy. Electrolyte supplementation must be monitored frequently, as adjustment of supplementation rates may be required. An animal that appears hyperkalemic at the time of initial examination may become hypokalemic shortly after fluid therapy has begun. Hypokalemia can be treated by administering potassium as an IV continuous rate infusion (CRI) at a rate that should not exceed 0.5 mEq/kg/hour (Table 1).

Table 1. Potassium supplementation in hypokalemic animals.*

Serum potassium concentration

Potassium (mEq) added to
250 ml fluid bag









* Not to exceed 0.5 mEq/kg/hour without electrocardiographic monitoring

If higher doses are required, continuous electrocardiographic monitoring should be performed. Hypophosphatemia is corrected with an intravenous CRI of potassium phosphate (the solution contains 4.4 mEq/ml of potassium and 3 mM/ml of phosphate) at a rate of 0.03-0.12 mM phosphate/kg/hour. Administration of potassium must be taken into account when giving potassium phosphate for correction of hypophosphatemia. A magnesium sulfate solution (containing 4 mEq/ml) given IV as a CRI of 1 mEq/kg/24 hour has been used successfully for correction of hypomagnesemia. Toxicity of erroneously administered intravenous Mg has been reported in one diabetic cat and one dog with acute renal disease.13 Signs or Mg toxicity in these animals included vomiting, weakness, generalized flaccid muscle tone, mental dullness, bradycardia, respiratory depression, and hypotension.13 Care must be taken to administer IV magnesium only to patients that have documented decreased iMg concentration. If the hyponatremia and hypochloremia persist, they are corrected by administering a saline solution of 0.9% NaCl.

Correction of hyperglycemia is performed by administering a rapidly acting insulin. Although several new rapidly acting insulin products have been introduced to the market and are being used successfully in management of humans with DKA, their clinical use in dogs and cats with DKA has not been reported. Therefore, at the moment, the use of regular insulin is recommended (Humulin R®, Novolin R®). Regular insulin is administered as an IV CRI (Table 2)14 or intramuscularly (IM).15

Table 2. Administration of IV regular insulin in patients with DKA.*

Blood glucose concentration

Fluid composition

Rate of administration


0.9% NaCl



0.45% NaCl + 2.5% dextrose



0.45% NaCl + 2.5% dextrose



0.45% NaCl + 5% dextrose



0.45% NaCl + 5% dextrose

Stop fluid administration

* 2.2 U/kg of regular crystalline insulin added to 250 ml of 0.9 % NaCl solution. The administration set must be flushed with 50 mL of the mixture prior to administration of the solution to the animal.

When IV regular insulin is administered as a CRI, blood glucose is measured every 2 hours. When insulin is administered IM, it is given every hour, and blood glucose is measured every hour. The initial dose of IM therapy is 0.2 U/kg regular insulin IM, followed by 0.1 U/kg regular insulin IM 1 hour later. Treatment with IM regular insulin is continued with 0.05 U/kg/hr, 0.1 U/kg/hr, or 0.2 U/kg/hr if blood glucose drops more by than 75 mg/dl/hour, between 50-75 mg/dl/hour, or by less than 50 mg/dl/hour, respectively.7

Acidosis is usually corrected with IV fluid administration and insulin therapy alone.2,4,12 The use of bicarbonate treatment for correction of acidosis in humans with DKA is controversial.12,16,17,18 The American Diabetes Association recommends bicarbonate supplementation only in DKA patients in which arterial pH remains less than 7.0 after 1 hour of fluid therapy.6 Possible risks associated with bicarbonate treatment in humans with DKA include exacerbation of hypokalemia, increased hepatic production of ketones, paradoxical cerebrospinal fluid acidosis, cerebral edema, and worsening intracellular acidosis due to increased production of carbon dioxide.2,17,18 Bicarbonate treatment is not needed in most dogs and cats with DKA.2,4 However, a recent retrospective study of 127 dogs with DKA reported that the degree of acidosis was associated with poor outcome.2 The same study reported that IV sodium bicarbonate therapy was associated with poor outcome.2 It is not known if bicarbonate therapy in itself, or the severe degree of acidosis that prompted such therapy was the cause of poor outcome in dogs treated with bicarbonate. A possible bicarbonate treatment protocol is to administer sodium bicarbonate at 1/2 to 1/3 of [0.3 X body weight X negative base excess] over a 20 minute interval, every 1 hour, while monitoring venous pH every hour. However, there are no studies to support this or any other specific bicarbonate treatment protocol in dogs and cats with DKA. The American Diabetes Association recommends treating pediatric DKA patients that maintain a pH of less than 7.0 after one hour of fluid therapy with 2 mEq/kg sodium bicarbonate added to 0.9% NaCl in a solution that does not exceed 155 mEq/L of sodium, over 1 hour.16 The pH is monitored every 1 hour and treatment is repeated until pH is 7.0 or greater.16

Presence of concurrent disease is believed to contribute to development of DKA. Therefore, identification of concurrent disease and specific treatment directed at alleviating any concurrent disease is indicated. It is possible that treatment of concurrent disease decreases secretion of glucagon and contributes to improved diabetic regulation and resolution of DKA.


Most dogs and cats (70%) treated for DKA survive to be discharged from the hospital.2,4 Median hospitalization time for dogs and cats with DKA is 6 and 5 days, respectively.2,4 At least 7% of dogs and up to 40% of cats develop recurring episodes of DKA.2,4 Dogs with coexisting hyperadrenocorticism are less likely to be discharged from the hospital, and the degree of base deficit in dogs is associated with outcome.2


1.  Ganong WF. 1997, Appelton & Lange.

2.  Hume DZ, et al. JVIM 2006 ;20:547-555

3.  Parsons SE, et al. JVECC 2002;12:147-152.

4.  Bruskiewicz KA, et al. JAVMA 1997;211:188

5.  Crenshaw KL, et al. JAVMA 1996 ;209:943.

6.  Christopher MM, et al. JAVMA 1995:9:24,

7.  Feldman EC, et al. Canine and Feline Endocrinology and Reproduction. Philadelphia, 2004,

8.  Willard MD, et al. JAVMA 1987:190:1007,

9.  Norris CR, et al. JAVMA 1999:215:1455,

10. Fincham SC, et al. JVIM 2004 :18:612,

11. Duarte R, et al. JVIM2002:16: 411,

12. Glaser N, et al. Ped Emerg Care2004:20:477,

13. Jackson CB, et al. JVECC 2004:14:115,

14. Macintire DK. 1993:JAVMA 202:1266,

15. Chastain CB, et al. JAVMA1981 :178:561,

16. American Diabetes Association 2003:26:S109,

17. Chiasson JL, et al. CMAJ2003 :168:859,

18. Okuda Y, et al. J Clin Endocrinol Metab 1996:81:314.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Rebecka Hess, DVM, DACVIM
University of Pennsylvania
Philadelphia, PA

MAIN : ACVIM SAIM Generalist : Diabetic Ketoacidosis
Powered By VIN