The Worst Diabetic Crisis

Holly Ahlgrim, DVM, walks us through Hyperglycemic Hyperosmolar Syndrome (HHS), discussing the signs, symptoms, and treatments for this condition.

Views: 7259 - Comments: 2

You are here

Diabetes mellitus is arguably one of the most common and frustrating endocrine diseases to treat in both human and veterinary medicine. It seems that no two patients are exactly alike in the way they present, the laboratory findings, and the way they respond to insulin therapy. Diabetics that are well regulated have a good long-term prognosis and can live relatively normal lives. However, diabetes mellitus (DM) becomes more complicated when either insulin therapy is inadequate or other concurrent illnesses occur, leading to decompensation. While there are many potential short and long-term complications of inadequately regulated DM, the two most serious sequelae are, without a doubt, diabetic ketoacidosis (DKA) and hyperglycemic hyperosmolar syndrome (HHS). While both of these conditions are potentially fatal, HHS carries a far more guarded prognosis with a lower survival rate. The focus of this article is a review of HHS.

HHS was, until recently, known as hyperosmolar non-ketotic coma or hyperosmolar non-ketotic diabetes mellitus in human literature. This terminology has been changed since coma, while a common presenting complaint in humans, is still inconsistently part of the presentation of this syndrome. The definition has also been broadened to exclude non-ketotic, since it is not uncommon for these patients to also have small to moderate ketonemia and ketonuria. Currently, the literature consensus is that HHS exists in patients with extreme hyperglycemia (>600 mg/dL) and severely elevated serum osmolality (>350 mOsm/L).


For HHS to develop, several things occur simultaneously. Similar to DKA, HHS develops due to a relative or absolute deficiency of insulin. HHS patients may be newly diagnosed diabetics, though many HHS patients are known diabetics. An insulin deficiency can develop for several reasons in known diabetics: missed doses of insulin, expired or spoiled insulin, or the development of another disease process. The development of HHS involves several counter regulatory hormones. These hormones (glucagon, epinephrine, cortisol and growth hormone) are present in the body at elevated levels due to concurrent illness. They all counteract the effects of insulin in different ways and lead to exacerbation of hyperglycemia. Ketogenesis is uncommon in HHS patients since the amount of insulin required to prevent ketone formation is approximately 10 times less compared to that required to significantly impact blood glucose concentrations. Additionally, for the degree of hyperglycemia to develop in HHS, a drop in glomerular filtration (GFR) must also occur. Decreased GFR can occur for many reasons. In animals that have preexisting renal disease, GFR can already be suboptimal. Hypovolemia from vomiting, decreased water intake, and progressively worsening osmotic diuresis all contribute to a decrease in GFR.

Neurologic signs develop due to the severe hyperosmolality in these patients. Hyperosmolality usually develops over days to weeks and induces the formation of idiogenic osmoles in the brain. These osmoles guard against dehydration of cerebral tissue. These osmotically active particles take time to form and take time to be eliminated. As fluid deficits are corrected, if osmolality drops too quickly in the serum, cerebral edema can ensue.

Because counter-regulatory hormones play a large role in the development of HHS, concurrent illnesses are nearly always present in diabetics that develop HHS. The most common concurrent illnesses reported in cats include renal failure, hyperthyroidism and acromegaly. The most common concurrent illnesses in dogs are reported to be pancreatitis and hyperadrenocorticism.

Physical Exam Findings

Patients with HHS always present with clinically severe dehydration (often >10 percent) due to osmotic diuresis from extreme hyperglycemia. HHS patients may present with altered or dull mentation and, in extreme cases, coma. Physical exam findings otherwise vary widely with any underlying disease processes. Those with concurrent renal disease may have small or enlarged kidneys, oral ulcerations and uremic breath. Patients with concurrent cardiac disease may show signs of congestive failure, respiratory distress, low body temperature and poor pulse quality. Patients with pancreatitis may exhibit signs of abdominal pain and nausea.


Any ill patient with a serum glucose of >600 mg/dL should be suspected to have HHS. Serum osmolality can be easily calculated with information gained from serum chemistries:

Serum Osm(calc) = 2(Na + K) + BUN/2.8 + glucose/18

Alternatively, since sodium is the number one contributor to osmolality, effective osmolality may also be used:

Effective Osm = 2(Na) + glucose/18

The most common concurrent laboratory finding in HHS is azotemia. Due to the inherent drop in GFR required to create HHS, almost all HHS patients are azotemic to some degree. Azotemia can be both pre-renal and renal in origin. Hyperphosphatemia often accompanies azotemia, though this frequently drops once insulin therapy is initiated.

Electrolyte derangements can vary. Patients are often initially hyponatremic due to the degree of hyperglycemia. Fluid is osmotically pulled from the interstitial space and falsely lowers sodium. HHS patients are actually usually hypernatremic due to massive free water loss. Calculating corrected sodium can help give a more accurate representation of the actual sodium level:

Na(corrected) = 1.6 x ([measured glucose – normal glucose]/100) + Na(measured)

It is also common for HHS patients to present with normal to increased potassium levels on serum chemistries. It is important to remember that despite this, these patients are actually potassium depleted. Intracellular potassium exchanges with hydrogen ions in states of acidosis to act as a physiologic buffer. As pH is corrected and insulin therapy is initiated, potassium moves back into the intracellular space and serum levels can drop precipitously. It is important to monitor electrolytes closely and intervene with supplementation to prevent severe hypokalemia.

While HHS patients are not as commonly acidotic as patients in DKA, it is still prudent to assess the blood gas status in these cases. Abnormal findings will generally include metabolic acidosis with a compensatory respiratory alkalosis. Acidemia in HHS patients is due to uremia and lactic acidosis, as opposed to ketoacidosis in DKA patients. Because of this, acid base disturbances often correct rapidly once fluid therapy is initiated.

In addition to identification of HHS in a patient, it is just as important to screen for concurrent illnesses. The knowledge of the presence of these conditions is important for both treatment and prognostication. In addition to a minimum database, abdominal ultrasound, thoracic radiographs, thyroid levels (particularly in cats), urine culture and species specific pancreatic lipase tests should be considered based on patient presentation.


The first step in treating these patients should always be volume resuscitation prior to starting insulin therapy. Volume resuscitation will drop serum glucose levels via an increase in GFR and simple dilution. Goals should be to replace volume deficits over 12 to 36 hours depending on the patient. Patients with concurrent cardiac or renal disease require more careful and slow fluid replacement to avoid fluid overload.

Replacing fluid deficits too fast can also lead to cerebral edema. Therefore, careful monitoring of electrolytes and the rate of drop in blood glucose is vital. Though sodium levels are frequently initially low, once glucose levels begin to drop, sodium will usually begin to rise. Because sodium is the number one contributor to serum osmolality, hypernatremia should not be corrected too quickly. A drop in serum sodium levels of no more than 1 mEq/L per hour is recommended. The type of crystalloid solution used often needs to be adjusted once, if not multiple times per day.

Institution of insulin therapy should be done with care. As previously mentioned, it is vital that blood glucose levels not drop too quickly to prevent rapid changes in osmolality. Due to the level of dehydration in these patients, initial insulin therapy should ideally be short acting insulin delivered through intravenous continuous rate infusion (CRI). This allows for easier adjustment of the amount of insulin delivered and is faster acting than subcutaneous or intramuscular administration. A CRI of regular or short acting insulin can be made by adding 1.1 U of regular insulin per kg for cats and 2.2 U/kg for dogs to 250 mL of 0.9 percent saline. Running this CRI at 10 mL/hr will deliver 1.1 U/kg/day for cats and 2.2 U/kg/day for dogs. Some advocate for a 50 percent reduction in the amount of insulin added to the CRI when treating HHS to avoid such a rapid drop in the blood glucose level. Since many patients are small, and the volume delivered with an insulin CRI can constitute a significant portion of the total crystalloid volume delivered, patient size should be taken into account when deciding how concentrated to make an insulin CRI. In general, a blood glucose drop of 50-75 mg/dL per hour should be the goal.

Acidosis rarely requires direct intervention. Because acidemia in HHS is due to uremia and lactic acidosis, IV fluid administration and improvement in GFR will generally indirectly correct any metabolic acidosis that is present. Treatment with sodium bicarbonate is usually unnecessary and is controversial, as it can worsen electrolyte disturbances.

Nutritional support is also important to consider in these patients to prevent further complications, particularly in obese cats. Enteral nutrition is always preferable to parenteral nutrition. Many patients may be unwilling to eat. Nasoesophageal or nasogastric tube placement should then be considered to initiate microenteral nutrition, working up to resting energy requirements. In patients with severe pancreatitis or gastrointestinal disease, enteral nutrition may not be well tolerated. Parenteral nutrition may have to be considered in patients that do not tolerate enteral nutrition. If parenteral nutrition is instituted, careful consideration should be taken on how best to balance the caloric distribution between protein, dextrose and lipids. Patients that have significant renal disease should be more sparingly protein fed. Those with pancreatitis may have issues tolerating a significant portion of calories from lipids. Inevitably, a portion of calories will come from dextrose. Because of this, higher levels of insulin supplementation will be necessary once parenteral nutrition is initiated. Long-term, esophagostomy tube placement can be considered, but is generally not considered for more critical hospitalized patients.

Lastly, it is important to treat concurrent illness. In patients with renal disease, GI disease and pancreatitis, fluid support is a large part of this. Analgesics, anti-emetics and gastro-protectants are commonly used in these patients as part of general supportive care. If there is evidence of urinary tract infection, pyelonephritis or other infectious processes, antibiotic therapy should be instituted. Conversely, in dogs suspected of having hyperadrenocorticism, treatment should not be initiated until the diagnosis can be confirmed with low-dose dexamethasone suppression testing. Because this test is significantly altered in illness, testing should be delayed until the HHS crisis is resolved and the dog has been discharged.


One of the most important things to help owners make decisions is prognosis. Unfortunately, HHS typically has a guarded to poor prognosis. Mortality rates in people with HHS are reported to be as high as 20 percent, roughly 10 times higher than the reported mortality rate for DKA. Veterinary patients with HHS are reported to have mortality rates ranging as high as approximately 40 to 60 percent in hospitals.


In conclusion, HHS is not as well defined in veterinary patients as it is in people. The amount of literature on HHS in dogs and cats is still relatively small compared to the amount of human medical literature. More studies are needed in veterinary medicine to further characterize this syndrome in dogs and cats. Ideally, studies should be directed at identifying the most common underlying disease processes and risk factors to aid in quicker identification of HHS. Early identification of HHS is important for both treatment and prognostication.


1. Schermerhorn T. Hyperosmolar non-ketotic syndrome in diabetic dogs. 25th ECVIM-CA Congress 2015.

2. Koenig A. Hyperglycemic Hyperosmolar Syndrome. IVECCS Proceedings 2010.

3. Rand J. Diabetic ketoacidosis and hyperosmolar hyperglycemic state in cats. Vet Clin Small Anim 2013;43(2):367-379.

 4. O’Brien MA. Diabetic emergencies in small animals. Vet Clin Small Anim 2010;40(2):317-333.

5. Trotman TK, Drobatz KJ, Hess RS. Retrospective evaluation of hyperosmolar hyperglycemia in 66 dogs (1993-2008). J Vet Emerg Crit Care 2013;23(5):557-564.

Sidebar Bookmark Button


Add To Training Plan


Content Assignment



charley abernathy's picture

So there must be a contraindication/binding effect to putting insulin in NormR? It seems NaCl is such an acidotic fluid for a DKA or even HHS and running it at that rate (0.45%+dextrose or 0.9%) doesn't allow for a lot of adjustment with other fluids for the Na for a cat, in particular...

Ladan Mohammad-Zadeh's picture

Hi! Thank you for your very good question. Yes, in fact, insulin is compatible with most crystalloid solutions, buffered isotonic varieties included (Normosol R, LRS, Plasmalyte). There are a few reasons why most resources list 0.9% saline as the base solution. NaCl tends to be less expensive compared to other crystalloid solutions. NaCl comes in a multitude of sizes, 100 ml, 250 ml, 500 ml etc, making the preparation of the CRI bag less cumbersome. The small bag sizes are advantageous as well because it gives the clinician the flexibility to change the concentration if needed from bag to bag. With as frequently as the CRI goes up and down in rate, it doesn't seem prudent to put it in a 1 liter maintenance fluid bag.
To answer your question about acidity of NaCl, it does not seem to be a contributing factor to slowing down a recovery from acidosis and it certainly does not make it worse. There's probably a few reasons for this. The acidosis that patients suffering from DKA or HHS have is a combination of ketoacidosis, uremic acidosis and lactic acidosis all of which improve with fluid therapy, ANY fluid therapy. And of course once insulin is started, the ketoacidosis will resolve too. There is evidence to suggest that large large volume infusions of 0.9% saline can cause an acidosis but it takes a LOT of fluid to see this effect.
Your observation about allowing for Na adjustments in a cat is a valid point . Some clinicians tend to double the concentration of insulin in the CRI on purpose so we can run it at a lower rate, so we have more flexibility in adjusting the maintenance fluid which runs alongside of the insulin CRI. I hope this answers your questions!