Dog cat

Diabetes Mellitus

A page about diabetes mellitus in cats describing cause, clinical signs, diagnosis and control.


Diabetes mellitus is a fairly common endocrine disorder of dogs and cats. It is commonly seen in middle aged to older pets.
Diabetes mellitus is characterized by a relative or absolute insulin deficiency. This results in an inability to transport glucose from the bloodstream into cells. Blood glucose concentrations are elevated. If the renal threshold is exceeded, glucose is excreted in the urine, leading to polyuria and polydipsia. In addition, glucose metabolism is impaired which leads to a deterioration in the general condition of the animal and, if untreated, to death.


Diabetes mellitus can be classified as follows: 

Type 1 
·         Occurs due to immune-mediated destruction of the insulin-producing (beta, ß) cells of the pancreas 
·         Results in total absence of insulin production, cases (i.e. all diabetic dogs) with this form of diabetes mellitus must be treated with insulin 
·         Rare in cats 

Type 2 
·         Cause not well defined 
·         Characterised by obesity (animals may no longer be overweight at the time of presentation), loss of tissue sensitivity to insulin, and deposition of amyloid in the pancreatic islets 
·         The pancreas may be producing none, too little, normal amounts, or even excessive amounts of insulin, but this insulin is failing to achieve its normal action 
·         Most common form of diabetes mellitus in cats 
·         Most cats with type 2 diabetes mellitus require insulin for successful treatment, at least initially. This may result in clinical remission. 
·         Clinical remission is a particular feature of diabetes mellitus in cats. This means that the clinical signs and need for insulin treatment in diabetic cats may be transient in nature. 
·         The pathological basis of this “transient diabetes mellitus” is yet to be fully determined. 
·         There is currently no way of predicting whether diabetes mellitus will be transient or permanent at the time of diagnosis. 

Type 3 
This is where diabetes mellitus occurs secondary to another disease, such as: 
– Hyperthyroidism in cats 
– Acromegaly 
– Drug treatment (e.g. corticosteroids, progestagens) 
– Extended use of these drugs may result in permanent diabetes mellitus 
– Generalised disease of the pancreas, such as that caused by an invasive tumour or chronic pancreatitis, if a sufficient proportion of beta cells are destroyed. 
– The underlying condition should be treated and, if necessary, the diabetes mellitus


·         In the 1960’s, estimates of diabetes were 1:800 in cats. In the 1980’s, the incidence was reported to be between 1:250 and 1:1,500 in cats. More recently the reported incidence is around 1:500 (0.2%) in both dogs and cats. 
·         Over one-half of diabetic cats are over 10 years old, and around three-quarters are 7 years old or older. 
·         In cats a higher incidence is seen in (neutered) males (1.5 males to 1 female). There is no known breed predilection in cats. Although workers in Australia noted that Burmese cats were over-represented in their studies.

Clinical Signs

In cats, the clinical signs are subtle and progressive, over several months. Early signs may include intermittent vomiting, anorexia and even diarrhoea. The first sign of diabetes mellitus may even be peripheral neuropathy. These symptoms are often seen before the onset of polydipsia (drinking a lot) and polyuria (urinating a lot). Only a small percentage of diabetic cats are polyphagic. 
The complications of diabetes mellitus that are seen in cats are: 

  • Diabetic ketoacidosis (DKA) 
    These animals have often been ill for only a short time. A variety of signs including severe depression, anorexia, vomiting, oliguria (reduced volume of urine) or anuria (not passing urine) and hyperventilation may be observed. These animals are often in shock and may have a distinct smell of acetone on their breath. 
    Hyperosmolar nonketotic syndrome: These animals have often been ill for longer than animals with DKA. Animals with hyperosmolar syndrome have extremely high blood glucose concentrations without ketoacidosis. These animals are usually very lethargic or comatose.
  • Diabetic neuropathy – a paralysis of limbs or other neurological effects


Diagnosis can be based on:

a) Clinical Signs

The clinical findings are non-specific. The physical findings may include 
·         Lethargy 
·         Dehydration 
·         Unkept hair coat and dehydration 

b) Laboratory tests 

A diagnosis of Diabetes Mellitus is based on the clinical signs and the presence of a persistent fasting hyperglycaemia and glucosuria. 
The normal reference range for blood glucose in cats is 3.4 to 5.7 mmol/l (61 to 103 mg/dl) 
Transient hyperglycaemia (10-15 mmol/l or 180-270 mg/dl) can occur in cats in stress situations, such as urinary tract infection or other concurrent disease, handling for diagnostic procedures (e.g., blood sampling). If there is doubt about the significance of the hyperglycaemia in a cat, the sampling should be repeated and assessed in combination with measurement of urine glucose and measurement of plasma fructosamine. 

The reference range for fructosamine varies from laboratory to laboratory, however the following can be used for guidance: 

·         In cats:  175-400 µmol/L 
Fructosamine concentrations <350 µmol/l rule out diabetes mellitus (persistent hyperglycaemia) in cats. 

The renal threshold for glucose is a blood glucose concentration of around: 

·         In cats: 14 mmol/l (250 mg/dl) 
If the blood glucose level exceeds this threshold, glucose is excreted in the urine.


The aim of insulin therapy in pets is to attain reasonable control of the clinical signs of diabetes mellitus and blood glucose concentrations by using once or twice daily injections of medium to long acting insulin preparations. A novel device for administering insulin to pets has been introduced to the Irish market by MSD Animal Health. Further information on the use of this device, called Vetpen, is available by clicking here.

As there is individual variation in the response to given insulin, an initial trial or stabilization period is usually undertaken. The initial stabilization period can last for up to 6 months. The owner should continue to monitor their animal at home, by either checking blood or urine glucose concentrations. As each animal’s insulin requirements change with time, periodic rechecks are essential, even after the initial stabilization period. These rechecks should include a clinical examination plus the construction of a blood glucose curve and some additional laboratory tests. 
If a cat is in crisis or in a coma, it should first be treated with fluids and intravenous insulin.  
It should be understood that it is impossible to maintain a normal glucose concentration throughout the entire day. If the blood glucose concentration can be maintained between 5 and 15 mmol/l (90 and 270 mg/dl) during a substantial part of the day, the clinical signs disappear and the main aims of therapy will have been achieved. 

Dose adjustment 
Dose adjustment should be based on biochemical assessment of the glycaemia. There are two body fluids that can be sampled to assess glycaemic control. 
·         Urine 
·         Blood 

Urine glucose monitoring 
Many vets recommend urine glucose testing as a method of monitoring diabetes mellitus in cats. This method is simple and inexpensive but it has some serious limitations that must be understood and taken into consideration.

This is based on the fact that once the blood glucose concentration exceeds the renal threshold; glucose is spilled into the urine. The amount of glucose present in the urine depends on how high the blood glucose was, and how long the blood glucose was high. In general, urine testing is more useful in dogs than in cats. 
Blood glucose monitoring 
This is a way of testing how much glucose is in the blood. There are a number of ways of assessing current blood glucose concentrations. 
·         Collecting a blood sample in a tube containing fluoride as an anticoagulant and submitting this to a laboratory for analysis. Plasma glucose concentrations are measured (“gold standard”). 
·         Collecting a drop of blood from the ear (pinna) or a carpal pad or a footpad and analyzing this using a hand-held blood glucose meter. A small test strip with the drop of blood on it is by inserted into a small machine (hand-held blood glucose meter), which reads the strip and shows the blood glucose level in a digital display window. The drop of blood can be produced using a sterile needle or a special lancet (razor sharp device to puncture the skin). 
·         Most monitors measure the glucose concentration in whole blood (not plasma) 
§  The readings may vary by as much as 15% from samples submitted to the laboratory. 
§   Hand-held meters are reasonably accurate. If a reading seems unusual or does not match the clinical signs, a second reading should be taken or another method used to measure blood glucose. 
§  All blood glucose machines are least accurate during episodes of hypoglycemia (low blood sugar). 

·         Collecting a drop of blood from the ear (pinna) or a carpal pad or a footpad and analyzing this using a test strip. The test strip changes colour according to how much glucose is in the blood. After a short incubation period the colour change on the test strip should be compared with the colour chart on the accompanying packaging. 
·         A number of new methods are being tested. These include continuous, less invasive methods of glucose testing such as using a glucose sensor that can be placed under the skin that can record several hundred blood glucose levels over a two- to three-day period. This type of device is currently being tested in diabetic cats. 
 The results of these tests should be recorded and kept in a record book or on a chart. 

Blood glucose curves 

This is a series of blood glucose measurements made after a dose of insulin is administered. It is the most effective way to monitor insulin therapy and the only way to determine how the insulin is working in a given individual. Assumptions about the time to peak insulin action and the duration of action are often inaccurate. 
To make a standard glucose curve, blood samples are taken just before and every 1.5 to 2 hours after insulin administration ideally until the next insulin dose is administered. 
This is not always possible and an alternative is to take blood samples every 1.5-2 hours until the nadir blood glucose concentration has been reached. After this a further one or two blood samples should be taken at 1.5-2 hour intervals to make sure the glucose concentrations are increasing and to determine the point that the renal threshold is passed (duration of action). This is best done by plotting the results as a graph.

This is only necessary if it is indicated i.e. not every occasion in every animal if they are doing well, drinking normally, and gaining weight and appear in general well controlled. Every 3-6 months would be a suitable interval. 
It is important that the blood samples always be taken from the same site e.g. capillary blood (pinna), cephalic/femoral vein or jugular vein. 
This will give information about the: 
·         onset of insulin action 
·         time of peak insulin action (time of nadir blood glucose) 
·         suitability of the dose (nadir blood glucose concentration) 
·         duration of insulin action 
It should be remembered that the time of peak insulin action may vary, e.g. if there is insulin resistance during early stabilization this may be delayed, if too high a dose is given the peak action may be earlier. It is thus important to repeat a blood glucose curve once the initial stabilization period is complete. 
The majority of cats receive insulin twice daily. It is not always necessary o perform a 24 hour glucose curve, since it has been shown that, at least in cats, the nadir concentration at night is likely to be slightly higher than during the day. Thus, there is less risk of hypoglycaemia developing at night. 
A “mini-glucose curve” is used to identify the highest and lowest, or peak and trough, blood glucose concentrations. This can be used in some situations instead of a full glucose curve. Normally the samples are taken just before insulin treatment (highest concentration) and at the time of the expected peak insulin action (based on what is known from a full glucose curve). 
Single measurements taken at the time when the insulin action is thought to peak are no longer considered helpful. 
1.     To establish initial insulin protocol at the time of diagnosis    
2.     Dose 
2.     To monitor the degree of regulation 
3.     To assess problems with regulation; If the response to insulin therapy is poor, a plasma glucose curve should be plotted, and every diagnostic effort should be made to rule out other problems. 
4.     To rule in or out rebound hyperglycemia 
This occurs when nadir blood glucose concentrations are <3.6 mmol/l (65 mg/dl). In response, the blood glucose concentration rises rapidly and may exceed 33 mmol/l (600 mg/dl), which appears as a HIGH reading on most blood glucose meters. This requires a full glucose curve. 

There are a number of ways of assessing past blood glucose concentrations 
·         Fructosamine 
·         Glycosylated (glycated) haemoglobin 

Fructosamines are stable complexes of carbohydrates and proteins that are produced by an irreversible, nonenzymatic glycosylation of protein. Glucose has a greater affinity for globulins in cats. Fructosamine measurements can be used to diagnose and monitor diabetes mellitus in cats. A single measure of fructosamine indicates the average glucose concentration over the previous 1-2 weeks. 
·         Highly sensitive 
·         Specific 
·         Distinguishes hyperglycemic, non-diabetic patients from diabetics with chronic hyperglycemia 

·         Does not appear to be influenced by transient hyperglycaemia (stress hyperglycaemia in cats) 
·         Depends on the level and duration of serum glucose concentration and the rate of turnover of specific plasma proteins.   
·         Changes more rapidly than glycosylated haemoglobin concentrations 
·         Useful in evaluating longer-term control and owner compliance with the administration of insulin 
The determination of fructosamine concentration in the blood is performed using an adapted, commercially available, automated, colorimetric nitroblue tetrazolium technique. This laboratory test is fast, reproducible, inexpensive, easily automated and precise. 
Although fructosamine levels are useful for evaluating the overall control of diabetes mellitus and long-term glucose regulation, there are some limitations. 
·         Unable to detect short-term or transient abnormalities in blood glucose, e.g., transient daily episodes of hypoglycaemia and/or hyperglycemia. This requires serial blood glucose measurements 
·         Hyperthyroid cats may have decreased fructosamine levels, despite having normal serum protein levels due to an increase in the protein turnover rate (decreased protein half-life) due to increased thyroid hormone levels. 
·         Fructosamine increases in hyperproteinaemia (e.g. dehydration)  
·         Globulin and fructosamine concentrations are correlated in cats. Hypoglobulinemia will result in a decreased fructosamine concentration. 

Glycosylated haemoglobin 

Glycosylated haemoglobin is produced by the non-enzymatic, irreversible binding of glucose to haemoglobin in erythrocytes. As the blood glucose concentration increases, the rate of hemoglobin glycosylation also increases. The specific fraction of glycosylated hemoglobin that is measured is HbA1c. This is reported as a percentage of hemoglobin that is in the glycosylated form. Glycosylated (glycated) hemoglobin concentration can be used as a screening test for diabetes mellitus, as well as for the monitoring of glycemic control in treated diabetic animals. 
Therefore, glycosylated hemoglobin is a viable alternative to fasting glucose measurements because it is unaffected by stress related or post-prandial hyperglycemia. Glycosylated hemoglobin determination also is useful in long term monitoring of diabetic patients over the previous 2-3 months. The binding of glucose and hemoglobin is irreversible, the affected senescent erythrocytes must be degraded before the glycosylated hemoglobin value will decrease. Glycosylated hemoglobin is not used routinely to monitor diabetic cats. 
·         Test not widely available for these species 
·         Not the most effective test due to the relatively long erythrocyte lifespan life span (approximately 70 days in cats). 
·         Less effective for short-term monitoring than fructosamine because hyperglycemia must be present for at least 3 weeks before increased HbA1c values are detectable 
·         Affected by haemoglobin concentrations: may be increased or decreased due to polycythemia or anemia (renal disease), respectively. 

Diet and exercise  

 Diet and exercise need to be controlled as these affect the insulin requirements. 
Dietary management is extremely important in the successful control of diabetes mellitus. Ideally the diet should be exactly the same every day and given at the same time and tit-bits should not be given to avoid fluctuations in blood glucose concentrations. If titbits are given, a slice of carrot is suitable – high fiber, low fat! Some general guidelines are described below. It is very important that the routine chosen should be: 
·         Suitable for the owner 
·         Minimise blood glucose fluctuations (post-prandial hyperglycaemia) 
In cats that are underweight, it is desirable that weight is gained (“ideal” body weight). 

Very calorie dense diets should be avoided, especially those that are high in soluble carbohydrates. In cats that are overweight, it is desirable that weight is lost (“ideal” body weight) in a controlled fashion. Most diets designed to promote weight loss are in fact high fibre diets. 
Water should be available at all times and its reduced consumption indicates successful treatment of the diabetes mellitus. 
The same amount of moderate exercise should be given every day. Wide variations in exercise will affect insulin requirements. 
Commercial diets produced for diabetic cats are available from veterinary surgeons. These diets are ideal for diabetic cats as they release glucose more slowly during the day and can result in a lower requirement for insulin. However, cats are known for their fussy eating habits and often find high fibre diets unpalatable. Anorexia is more dangerous that hyperglycaemia so in this situation the diabetic cat should be maintained on its usual diet (preferably exactly the same type of food every day). 
Meal feeding of cats may not be possible as many cats prefer to browse, eating many small “snacks” (somewhere in the range of 5-11) every day. In cats, there also appears to be far less of a connection between feeding and hyperglycaemia. Thus, the need to time meals around the insulin treatment is far less. Ideally, a cat should be kept on its usual feeding routine (e.g. food always available (ad libitum), meals/fresh food given twice or three times daily). Insulin should always be administered twice daily to cats. 
The exception is for cats that are obese. These cats should be given a diet designed for weight management (these are high fibre diets) and fed according to a strict regime until the cat reaches its target body weight. In some cats, weight loss may eliminate the need for insulin therapy (“clinical diabetic remission”).

Further information for owners is available on

Further information for veterinary practitioners is available on