Attacks of DKA can be prevented in those known to have diabetes to an extent by adherence to "sick day rules"; these are clear-cut instructions to person on how to treat themselves when unwell. Instructions include advice on how much extra insulin to take when sugar levels appear uncontrolled, an easily digestible diet rich in salt and carbohydrates, means to suppress fever and treat infection, and recommendations when to call for medical help.

People with diabetes can monitor their own ketone levels when unwell and seek help if they are elevated.

Children's blood sugar levels are often slightly lower than adults'. Overnight fasting glucose levels are below 70 mg/dL (3.9 mM) in 5% of healthy adults, but up to 5% of children can be below 60 mg/dL (3.3 mM) in the morning fasting state. As the duration of fasting is extended, a higher percentage of infants and children will have mildly low plasma glucose levels, usually without symptoms. The normal range of newborn blood sugars continues to be debated. It has been proposed that newborn brains are able to use alternate fuels when glucose levels are low more readily than adults. Experts continue to debate the significance and risk of such levels, though the trend has been to recommend maintenance of glucose levels above 60–70 mg/dL the first day after birth.

Diabetic hypoglycemia represents a special case with respect to the relationship of measured glucose and hypoglycemic symptoms for several reasons. First, although home glucose meter readings are often misleading, the probability that a low reading, whether accompanied by symptoms or not, represents real hypoglycemia is much higher in a person who takes insulin than in someone who does not.

Diabetic ketoacidosis may be diagnosed when the combination of hyperglycemia (high blood sugars), ketones in the blood or on urinalysis and acidosis are demonstrated. In about 10% of cases the blood sugar is not significantly elevated ("euglycemic diabetic ketoacidosis").

Arterial blood gas measurement is usually performed to demonstrate the acidosis; this requires taking a blood sample from an artery. Subsequent measurements (to ensure treatment is effective), may be taken from a normal blood test taken from a vein, as there is little difference between the arterial and the venous pH. Ketones can be measured in the urine (acetoacetate) and blood (β-hydroxybutyrate). When compared with urine acetoacetate testing, capillary blood β-hydroxybutyrate determination can reduce the need for admission, shorten the duration of hospital admission and potentially reduce the costs of hospital care. At very high levels, capillary blood ketone measurement becomes imprecise.

In addition to the above, blood samples are usually taken to measure urea and creatinine (measures of kidney function, which may be impaired in DKA as a result of dehydration) and electrolytes. Furthermore, markers of infection (complete blood count, C-reactive protein) and acute pancreatitis (amylase and lipase) may be measured. Given the need to exclude infection, chest radiography and urinalysis are usually performed.

If cerebral edema is suspected because of confusion, recurrent vomiting or other symptoms, computed tomography may be performed to assess its severity and to exclude other causes such as stroke.

Blood glucose levels discussed in this article are venous plasma or serum levels measured by standard, automated glucose oxidase methods used in medical laboratories. For clinical purposes, plasma and serum levels are similar enough to be interchangeable. Arterial plasma or serum levels are slightly higher than venous levels, and capillary levels are typically in between. This difference between arterial and venous levels is small in the fasting state but is amplified and can be greater than 10% in the postprandial state. On the other hand, whole blood glucose levels (e.g., by fingerprick meters) are about 10–15% lower than venous plasma levels. Furthermore, available fingerstick glucose meters are only warranted to be accurate to within 15% of a simultaneous laboratory value under optimal conditions, and home use in the investigation of hypoglycemia is fraught with misleading low numbers. In other words, a meter glucose reading of 39 mg/dL could be properly obtained from a person whose laboratory serum glucose was 53 mg/dL; even wider variations can occur with "real world" home use.

Two other factors significantly affect glucose measurement: hematocrit and delay after blood drawing. The disparity between venous and whole blood concentrations is greater when the hematocrit is high, as in newborn infants, or adults with polycythemia. High neonatal hematocrits are particularly likely to confound glucose measurement by meter. Second, unless the specimen is drawn into a fluoride tube or processed immediately to separate the serum or plasma from the cells, the measurable glucose will be gradually lowered by "in vitro" metabolism of the glucose at a rate of approximately 7 mg/dL/h, or even more in the presence of leukocytosis. The delay that occurs when blood is drawn at a satellite site and transported to a central laboratory hours later for routine processing is a common cause of mildly low glucose levels in general chemistry panels.

It is critical for patients who monitor glucose levels at home to be aware of which units of measurement their testing kit uses.

Glucose levels are measured in either:

1. Millimoles per liter (mmol/l) is the SI standard unit used in most countries around the world.

2. Milligrams per deciliter (mg/dl) is used in some countries such as the United States, Japan, France, Egypt and Colombia.

Scientific journals are moving towards using mmol/l; some journals now use mmol/l as the primary unit but quote mg/dl in parentheses.

Glucose levels vary before and after meals, and at various times of day; the definition of "normal" varies among medical professionals. In general, the normal range for most people (fasting adults) is about 4 to 6 mmol/l or 80 to 110 mg/dl. (where 4 mmol/l or 80 mg/dl is "optimal".) A subject with a consistent range above 7 mmol/l or 126 mg/dl is generally held to have hyperglycemia, whereas a consistent range below 4 mmol/l or 70 mg/dl is considered hypoglycemic. In fasting adults, blood plasma glucose should not exceed 7 mmol/l or 126 mg/dL. Sustained higher levels of blood sugar cause damage to the blood vessels and to the organs they supply, leading to the complications of diabetes.

Chronic hyperglycemia can be measured via the HbA1c test. The definition of acute hyperglycemia varies by study, with mmol/l levels from 8 to 15 (mg/dl levels from 144 to 270).

Defects in insulin secretion, insulin action, or both, results in hyperglycemia.

Cranial imaging is not used for diagnosis of this condition. However, if MRI is performed, it may show cortical restricted diffusion with unusual characteristics of reversible T2 hypointensity in the subcortical white matter.

Ketones in the urine or blood, as detected by urine strips or a blood ketone testing meter, may indicate the beginning of diabetic ketoacidosis (DKA), a dangerous and often quickly fatal condition caused by high glucose levels (hyperglycemia) and low insulin levels combined with certain other systemic stresses. DKA can be arrested if caught quickly.

Ketones are produced by the liver as part of fat metabolism and are normally not found in sufficient quantity to be measured in the urine or blood of non-diabetics or well-controlled diabetics. The body normally uses glucose as its fuel and is able to do so with sufficient insulin levels. When glucose is not available as an energy source because of untreated or poorly treated diabetes and some other unrelated medical conditions, it begins to use fat for energy instead. The result of the body turning to using fat instead of glucose for energy means ketone production that is measurable when testing either urine or blood for them.

Ketone problems that are more serious than the "trace or slight" range need immediate medical attention; they cannot be treated at home. Veterinary care for ketosis/ketoacidosis can involve intravenous (IV) fluids to counter dehydration, when necessary, to replace depleted electrolytes, intravenous or intramuscular short-acting insulin to lower blood glucose levels, and measured amounts of glucose or force feeding, to bring the metabolism back to using glucose instead of fat as its source of energy.

When testing urine for ketones, the sample needs to be as fresh as possible. Ketones evaporate quickly, so there is a chance of getting a false negative test result if testing older urine. The urine testing strip bottle has instructions and color charts to illustrate how the color on the strip will change given the level of ketones or glucose in the urine over 15 (ketones–Ketostix) or 30 (glucose–Ketodiastix) seconds. Reading the colors at those time intervals is important because the colors will continue to darken and a later reading will be an incorrect result. Timing with a clock or watch second hand instead of counting is more accurate.

At present, there is only one glucometer available for home use that tests blood for ketones using special strips for that purpose–Abbott's Precision Xtra. This meter is known as Precision, Optium, or Xceed outside of the US. The blood ketone test strips are very expensive; prices start at about US$50 for ten strips. It is most likely urine test strips–either ones that test only for ketones or ones that test for both glucose and ketones in urine would be used. The table above is a guide to when ketones may be present.

Insulin is given to reduce blood glucose concentration; however, as it also causes the movement of potassium into cells, serum potassium levels must be sufficiently high or dangerously low blood potassium levels may result. Once potassium levels have been verified to be greater than 3.3 mEq/l, then an insulin infusion of 0.1 units/kg/hr is started. The goal for resolution is a blood glucose of less than 200 mg/dL.

The guidelines for preventing impaired fasting glucose are the same as those given for preventing type 2 diabetes in general. If these are adhered to, the progression to clinical diabetes can be slowed or halted. In some cases, a complete reversal of IFG can be achieved. Certain risk factors, such as being of Afro-Caribbean or South Asian ethnicity, as well as increasing age, are unavoidable, and such individuals may be advised to follow these guidelines, as well as monitor their blood glucose levels, more closely.

Diabetic coma was a more significant diagnostic problem before the late 1970s, when glucose meters and rapid blood chemistry analyzers were not universally available in hospitals. In modern medical practice, it rarely takes more than a few questions, a quick look, and a glucose meter to determine the cause of unconsciousness in a patient with diabetes. Laboratory confirmation can usually be obtained in half an hour or less. Other conditions that can cause unconsciousness in a person with diabetes are stroke, uremic encephalopathy, alcohol, drug overdose, head injury, or seizure.

Fortunately, most episodes of diabetic hypoglycemia, DKA, and extreme hyperosmolarity do not reach unconsciousness before a family member or caretaker seeks medical help.

The use of an inexpensive glucometer and blood glucose testing at home can help avoid dangerous insulin overdoses and can provide a better picture of how well the condition is managed.

A 2003 study of canine diabetes caregivers who were new to testing blood glucose at home found 85% of them were able to both succeed at testing and to continue it on a long-term basis. Using only one blood glucose reading as the reason for an insulin dose increase is to be avoided; while the results may be higher than desired, further information, such as the lowest blood glucose reading or nadir, should be available to prevent possible hypoglycemia.

Urine strips are not recommended to be used as the sole factor for insulin adjustments as they are not accurate enough. Urine glucose testing strips have a negative result until the renal threshold of 10 mmol/L or 180 mg/dL is reached or exceeded for a period of time. The range of negative reading values is quite wide-covering normal or close to normal blood glucose values with no danger of hypoglycemia (euglycemia) to low blood glucose values (hypoglycemia) where treatment would be necessary. Because urine is normally retained in the bladder for a number of hours, the results of urine testing are not an accurate measurement of the levels of glucose in the bloodstream at the time of testing.

Glucometers made for humans are generally accurate using canine and feline blood except when reading lower ranges of blood glucose (<80 mg/dL), (<4.44 mmol/L). It is at this point where the size difference in human vs animal red blood cells can create inaccurate readings. Glucometers for humans were successfully used with pets long before animal-oriented meters were produced. A 2009 study directly compared readings from both types of glucometers to those of a chemistry analyzer. Neither glucometer's readings exactly matched those of the analyzer, but the differences of both were not clinically significant when compared to analyzer results. All glucometer readings need to be compared to same sample laboratory values to determine accuracy.

Nonketotic hyperosmolar coma usually develops more insidiously than DKA because the principal symptom is lethargy progressing to obtundation, rather than vomiting and an obvious illness. Extremely high blood sugar levels are accompanied by dehydration due to inadequate fluid intake. Coma from NKHC occurs most often in patients who develop type 2 or steroid diabetes and have an impaired ability to recognize thirst and drink. It is classically a nursing home condition but can occur in all ages.

The diagnosis is usually discovered when a chemistry screen performed because of obtundation reveals an extremely high blood sugar level (often above 1800 mg/dl (100 mM)) and dehydration. The treatment consists of insulin and gradual rehydration with intravenous fluids.

Fasting plasma glucose screening should begin at age 30–45 and be repeated at least every three years. Earlier and more frequent screening should be conducted in at-risk individuals. The risk factors for which are listed below:

- Family history (parent or sibling)

- Dyslipidemia (triglycerides > 200 or HDL < 35)

- Overweight or obesity (body mass index > 25)

- History of gestational diabetes or infant born with birth weight greater than

- High risk ethnic group

- Hypertension (systolic blood pressure >140 mmHg or diastolic blood pressure > 90 mmHg)

- Prior fasting blood glucose > 99

- Known vascular disease

- Markers of insulin resistance (PCOS, acanthosis nigricans)

Different organisations use slightly differing levels before classifying a person's fasting blood glucose as "impaired", with the American Diabetes Association using a lower cutoff in its criteria than the World Health Organization. The upper limits remain the same, as fasting levels above this are almost universally accepted as indicative of full diabetes:

- WHO criteria: fasting plasma glucose level from 6.1 mmol/l (110 mg/dL) to 6.9 mmol/l (125 mg/dL).

- ADA criteria: fasting plasma glucose level from 5.6 mmol/L (100 mg/dL) to 6.9 mmol/L (125 mg/dL).

The diagnosis is based on a combination of typical clinical features and exclusion by a pediatric endocrinologist of other causes of "hypoglycemia with ketosis," especially growth hormone deficiency, hypopituitarism, adrenal insufficiency, and identifiable inborn errors of metabolism such as organic acidoses.

The most useful diagnostic tests include measurement of insulin, growth hormone, cortisol, and lactic acid at the time of the hypoglycemia. Plasma acylcarnitine levels and urine organic acids exclude some of the important metabolic diseases. When the episodes are recurrent or severe, the definitive test is a hospitalization for a supervised diagnostic fast. This usually demonstrates "accelerated fasting"—a shorter time until the glucose begins to fall, but normal metabolic and counterregulatory responses as the glucose falls. As the glucose reaches hypoglycemic levels, the insulin is undetectable, counterregulatory hormones, fatty acids, and ketones are high, and glucagon injection elicits no rise of glucose.

If a person cannot receive oral glucose gel or tablets, such as the case with unconsciousness, seizures, or altered mental status, then emergency personnel (EMTs/Paramedics and in-hospital personnel) can establish a peripheral or central IV line and administer a solution containing dextrose and saline. These are normally referred to as Dextrose (Concentration) Water, and come in 5%, 10%, 25% and 50%. Dextrose 5% and 10% come in IV bag and syringe form, and are mainly used in infants and to provide a fluid medium for medications. Dextrose 25% and 50% are heavily necrotic due to their hyperosmolarity, and should only be given through a patent IV line - Any infiltration can cause massive tissue necrosis. CAUTION: Dextrose 25% and 50% can easily cause necrosis in small veins. It is MUCH safer to use a Dextrose 10% solution when treating hypoglycemia via IV in children under the age of 14. When using Dextrose 25% in a child it is safer to administer it through a central line or an intra-oseous line.

The blood glucose can usually be raised to normal within minutes with 15-20 grams of carbohydrate, although overtreatment should be avoided if at all possible. It can be taken as food or drink if the person is conscious and able to swallow. This amount of carbohydrate is contained in about 3-4 ounces (100-120 mL) of orange, apple, or grape juice, about 4-5 ounces (120-150 mL) of regular (non-diet) soda, about one slice of bread, about 4 crackers, or about 1 serving of most starchy foods. Starch is quickly digested to glucose, but adding fat or protein retards digestion. Composition of the treatment should be considered, as fruit juice is typically higher in fructose which takes the body longer to metabolize than simple dextrose alone. Following treatment, symptoms should begin to improve within 5 to 10 minutes, although full recovery may take 10–20 minutes. It should be noted that over treatment does not speed recovery, and will simply produce hyperglycemia afterwards, which ultimately will need to be corrected. On the other hand, since the excess of insulin over the amount required to normalize blood sugar may continue to reduce blood sugar levels after treatment has produced an initial normalization, continued monitoring is required to determine if further treatment is necessary.

The American College of Endocrinology (ACE) and the American Association of Clinical Endocrinologists (AACE) have developed "lifestyle intervention" guidelines for preventing the onset of type 2 diabetes:

- Healthy meals (a diet with no saturated and trans fats, sugars, and refined carbohydrates, as well as limited the intake of sodium and total calories)

- Physical exercise (30–45 minutes of cardio vascular exercise per day, five days a week)

- Reducing weight by as little as 5–10 percent may have a significant impact on overall health

Opinions differ about optimal screening and diagnostic measures, in part due to differences in population risks, cost-effectiveness considerations, and lack of an evidence base to support large national screening programs. The most elaborate regimen entails a random blood glucose test during a booking visit, a screening glucose challenge test around 24–28 weeks' gestation, followed by an OGTT if the tests are outside normal limits. If there is a high suspicion, a woman may be tested earlier.

In the United States, most obstetricians prefer universal screening with a screening glucose challenge test. In the United Kingdom, obstetric units often rely on risk factors and a random blood glucose test. The American Diabetes Association and the Society of Obstetricians and Gynaecologists of Canada recommend routine screening unless the woman is low risk (this means the woman must be younger than 25 years and have a body mass index less than 27, with no personal, ethnic or family risk factors) The Canadian Diabetes Association and the American College of Obstetricians and Gynecologists recommend universal screening. The U.S. Preventive Services Task Force found there is insufficient evidence to recommend for or against routine screening.

Some pregnant women and careproviders choose to forgo routine screening due to the absence of risk factors, however this is not advised due to the large proportion of women who develop gestational diabetes despite having no risk factors present and the dangers to the mother and baby if gestational diabetes remains untreated.

Absolute numbers vary between pets, and with meter calibrations. Glucometers made for humans are generally accurate using feline blood except when reading lower ranges of blood glucose (<80 mg/dl–4.44 mmol/L). At this point the size difference in human and animal red blood cells can create inaccurate readings.

No major organization recommends universal screening for diabetes as there is no evidence that such a program improve outcomes. Screening is recommended by the United States Preventive Services Task Force (USPSTF) in adults without symptoms whose blood pressure is greater than 135/80 mmHg. For those whose blood pressure is less, the evidence is insufficient to recommend for or against screening. There is no evidence that it changes the risk of death in this group of people. They also recommend screening among those who are overweight and between the ages of 40 and 70.

The World Health Organization recommends testing those groups at high risk and in 2014 the USPSTF is considering a similar recommendation. High-risk groups in the United States include: those over 45 years old; those with a first degree relative with diabetes; some ethnic groups, including Hispanics, African-Americans, and Native-Americans; a history of gestational diabetes; polycystic ovary syndrome; excess weight; and conditions associated with metabolic syndrome. The American Diabetes Association recommends screening those who have a BMI over 25 (in people of Asian descent screening is recommended for a BMI over 23).

Treatment of hyperglycemia requires elimination of the underlying cause, such as diabetes. Acute hyperglycemia can be treated by direct administration of insulin in most cases. Severe hyperglycemia can be treated with oral hypoglycemic therapy and lifestyle modification.

In diabetes mellitus (by far the most common cause of chronic hyperglycemia), treatment aims at maintaining blood glucose at a level as close to normal as possible, in order to avoid these serious long-term complications. This is done by a combination of proper diet, regular exercise, and insulin or other medication such as metformin, etc.

Those with hyperglycaemia can be treated using sulphonylureas or metformin or both. These drugs help by improving glycaemic control

Dipeptidyl peptidase 4 inhibitor alone or in combination with basal insulin can be used as a treatment for hyperglycemia with patients still in the hospital.

Children "outgrow" ketotic hypoglycemia, presumably because fasting tolerance improves as body mass increases. In most the episodes become milder and more infrequent by 4 to 5 years of age and rarely occur after age 9. Onset of hypoglycemia with ketosis after age 5 or persistence after age 7 should elicit referral and an intensive search for a more specific disease.

Women with GDM may have high glucose levels in their urine (glucosuria). Although dipstick testing is widely practiced, it performs poorly, and discontinuing routine dipstick testing has not been shown to cause underdiagnosis where universal screening is performed. Increased glomerular filtration rates during pregnancy contribute to some 50% of women having glucose in their urine on dipstick tests at some point during their pregnancy. The sensitivity of glucosuria for GDM in the first 2 trimesters is only around 10% and the positive predictive value is around 20%.

Diagnosis can be made by checking fasting and post prandial insulin levels either with normal meal or with 100gms of oral glucose