Epidemiology
Among adults, two thirds of episodes of diabetic ketoacidosis (DKA) occur in people with type 1 diabetes, while one third occur in those with type 2 diabetes.[11] Between 6% and 21% of adults with type 1 diabetes present with DKA at their initial diagnosis.[1] In people with pre-existing type 2 diabetes, DKA may be triggered by conditions of physiological stress (such as infections, trauma, or cardiovascular events), or by the use of certain drugs (e.g., sodium-glucose cotransporter-2 [SGLT2] inhibitors and the dual SGLT1/SGLT2 inhibitor sotagliflozin). DKA has also been increasingly documented as a presenting feature of newly diagnosed type 2 diabetes; this is referred to as ketosis-prone diabetes.[11][12] Since the early 2000s, the prevalence of ketosis-prone diabetes worldwide has increased, with epidemiological data suggesting that people of African, Hispanic, and Asian origin are at greater risk.[11][13] Most often, individuals with ketosis-prone diabetes have obesity and a strong family history of type 2 diabetes, as well as evidence of insulin resistance.[11]
As the majority of people with DKA are hospitalised, most epidemiological data come from hospital discharge coding.[11] In England, the incidence of hospital admissions for DKA among adults with type 2 diabetes increased by 4% annually between 1998 and 2013; hospitalisations for DKA in adults with type 1 diabetes increased from 1998 to 2007, and remained static until 2013.[14]
In the US, from 2000 to 2009, the rate of hospitalisations for DKA decreased overall, from 21.9 to 19.5 per 1000 persons with diabetes, but then increased in the period 2009 to 2014 to 30.2 per 1000 persons with diabetes.[15] In 2014, rates of hospitalisation for DKA were highest among people aged <45 years (44.3 per 1000 persons with diabetes), and decreased with age (5.2 per 1000 persons with diabetes aged 45 to 64 years; 1.6 per 1000 persons aged 65 to 74 years; and 1.4 per 1000 persons aged ≥75 years).[15] The American Diabetes Association has raised concerns about the rise in the rate of hyperglycaemic crises in people with both type 1 and type 2 diabetes over the past decade, with data from 2020 suggesting hyperglycaemic crisis rates of 44.5 per 1000 person-years among people with type 1 diabetes and up to 3.2 per 1000 person-years among people with type 2 diabetes.[6][16] According to the US Centers for Disease Control and Prevention, 10.2 in 1000 adults with diabetes in the US visited emergency departments with DKA in 2020.[17]
During the period 2000 to 2014, in-hospital mortality rates among people with DKA consistently decreased in the US, from 1.1% to 0.4%.[15] Mortality rates reported in low- and middle-income countries are much higher, potentially because of delayed diagnosis and treatment.[1] Data from India have shown a 30% mortality rate in those presenting with DKA, and studies from sub-Saharan Africa have reported similarly high mortality (26% to 41%).[11]
Risk factors
Diabetic ketoacidosis can occur in people with both type 1 and type 2 diabetes, but it is much more common in those with type 1 diabetes. One US nationwide cohort study found that adjusted rates of hyperglycaemic crises were 53 events per 1000 person-years among adults with type 1 diabetes, compared with 4 events per 1000 person-years among those with type 2 diabetes.[34]
Reduction in the net effective concentration of insulin leads to impaired carbohydrate, lipid, and ketone metabolism in diabetic ketoacidosis (DKA). Decreased insulin results in increased gluconeogenesis, accelerated glycogenolysis, and impaired glucose utilisation by peripheral tissues.[1]
In one US study conducted in a large urban hospital, poor adherence to insulin treatment accounted for >50% of DKA admissions and was the principal cause for the development of DKA among African-Caribbean populations and underinsured people.[35] Psychological and social factors - such as mental health problems, poor health literacy, and lack of support - may negatively affect adherence to insulin therapy, particularly in disadvantaged populations.[36][37] Additionally, insulin pump failure (e.g., due to dislodgement or occlusion) can result in rapid DKA development.[6]
The most common precipitating factor for diabetic ketoacidosis (DKA) worldwide is infection, particularly urinary tract infection or pneumonia.[1][18] Increased levels of counter-regulatory hormones, particularly adrenaline, as part of the body’s systemic response to infection can lead to insulin resistance, increased lipolysis, ketogenesis, and volume depletion, all of which may contribute to the development of hyperglycaemic crises in patients with diabetes.[1] Coronavirus disease 2019 (COVID-19) infection has also been associated with greater risk of DKA in both type 1 and type 2 ԁiabеtеs.[1] In addition, the American Diabetes Association notes that several studies have reported DKA as the presenting feature of newly diagnosed type 1 diabetes during or following COVID-19 infection; however, the precise mechanisms underlying new-onset diabetes in people with COVID-19 remain unclear.[6]
Underlying cardiovascular events, particularly myocardial infarction, can trigger the release of counter-regulatory hormones such as catecholamines and cortisol, which promote insulin resistance, increase lipolysis, and enhance ketone body production, thereby predisposing patients with diabetes to diabetic ketoacidosis.[1][38]
A substantial proportion of individuals hospitalised with diabetic ketoacidosis (DKA) experience recurrent episodes.[6] In US nationwide studies, up to 22% of people admitted with DKA had at least one readmission within 30 days or during the same calendar year.[39] Among those readmitted within 30 days, 40.8% represented recurrent DKA episodes, with approximately 50% being readmitted within 2 weeks.[39][40] Among those readmitted within the same calendar year, 86% had 1 to 3 DKA readmissions and 14% had ≥4.[40] Careful assessment of precipitating and contributing factors, along with close follow-up within 2-4 weeks of discharge, may help reduce the risk of recurrence.[1]
A prior history of hypoglycaemic crises, indicative of greater glycaemic variability, is also a recognised risk factor for DKA.[1] One study reported that severe hypoglycaemia was associated with a three- to fourfold increase in the risk of subsequent hyperglycaemic crisis.[34]
Poor glycaemic control is strongly associated with an increased risk of hyperglycaemic crises.[6] One study found that, among patients with type 1 diabetes, the risk of experiencing a hyperglycaemic crisis increased when the HbA1c level exceeded 53 mmol/mol.[34] The incidence risk ratio was 7.81 (95% CI 5.78 to 10.54) for HbA1c levels of ≥86 mmol/mol compared with HbA1c levels of 48-52 mmol/mol.[34] In patients with type 2 diabetes, the risk increased progressively for all HbA1c levels above 38 mmol/mol, with an incidence risk ratio of 7.06 (95% CI 6.26 to 7.96) for HbA1c levels ≥86 mmol/mol.[34]
Adverse social determinants of health are among the strongest factors associated with recurrent diabetic ketoacidosis (DKA).[1] Multiple studies have demonstrated that low income, area-level deprivation, housing insecurity, and lack of insurance or presence of underinsurance (in health systems where this is applicable) are linked to an increased risk of DKA and hyperosmolar hyperglycaemic state, with approximately 40% of hyperglycaemic crises occurring in lower-income and underserved populations.[1]
One nationwide US cohort study found that rates of hyperglycaemic crises were significantly higher among younger adults with both type 1 and type 2 diabetes.[34] In 2014, rates of hospitalisation for diabetic ketoacidosis in the US were highest in people aged <45 years (44.3 per 1000 persons with diabetes) and declined with increasing age (5.2 per 1000 in those aged 45-64 years; 1.6 per 1000 in those aged 65-74 years; and 1.4 per 1000 in those aged ≥75 years).[9]
The presence of diabetic microvascular complications, including neuropathy, nephropathy, and retinopathy, has been associated with an increased risk of diabetic ketoacidosis (DKA).[34] In a large prospective cohort study, reduced estimated glomerular filtration rate (eGFR <60 mL/min/1.73 m²), even after excluding patients with end-stage renal disease, was linked to a higher risk of hospitalisation for DKA (HR 1.71, 95% CI 1.26 to 2.67) compared with individuals with eGFR >60 mL/min/1.73 m².[41] This association persisted independently of glycaemic control.
The presence of comorbidities - such as cerebrovascular disease, heart failure, dementia, chronic obstructive pulmonary disease, cirrhosis, or cancer - has been associated with an increased risk of diabetic ketoacidosis in patients with type 2 diabetes.[34]
Acute medical events such as stroke, characterised by increased levels of counter-regulatory hormones and compromised access to water and insulin, may contribute to the development of hyperglycaemic crises.[1]
Hormonal disturbances in certain endocrine disorders can increase counter-regulatory hormone levels, contributing to the development of diabetic ketoacidosis in patients with coexisting diabetes.[43]
Drugs that affect carbohydrate metabolism may precipitate hyperglycaemic crises.[1][19][20][22] This may include corticosteroids, thiazide diuretics, pentamidine, sympathomimetics (e.g., dobutamine, terbutaline), and atypical antipsychotics (e.g., clozapine, olanzapine, risperidone).[22] Cocaine use has been identified as an independent risk factor for recurrent diabetic ketoacidosis (DKA).[21] Cannabis use (and associated hyperemesis syndrome) has also been associated with an increased risk of DKA in adults with type 1 diabetes.[6]
Sodium-glucose cotransporter-2 (SGLT2) inhibitors (e.g., canagliflozin, dapagliflozin, empagliflozin, ertugliflozin), used for glycaemic control in type 2 diabetes and more recently for cardiorenal risk reduction, have been the subject of a US Food and Drug Administration (FDA) warning regarding DKA risk.[27] The risk is heightened in situations such as severe illness, prolonged fasting, or perioperative periods, and SGLT2 inhibitors should be avoided in these contexts.[6] The incidence of DKA is low, however, with only a modest incremental absolute risk; American Diabetes Association (ADA)/European Association for the Study of Diabetes (EASD) consensus guidelines note that the SGLT2 inhibitor cardiovascular outcome trials have reported DKA rates of 0.1% to 0.6% versus rates of <0.1% to 0.3% with placebo.[31] While SGLT2 inhibitors are associated with a low incidence of DKA in people with type 2 diabetes (0.6 to 4.9 events per 1000 patient-years), the ADA cautions that the risk of DKA can be substantially higher (5 to 17 times) in individuals with type 1 diabetes, affecting approximately 4% of patients.[6][46]
The dual SGLT1/SGLT2 inhibitor sotagliflozin has also been associated with increased rates of DKA.[6] In clinical trials of sotagliflozin in people with type 1 diabetes, results showed improvements in HbA1c and body weight, but use was associated with an eightfold increase in DKA compared with placebo.[6][47]
Immune checkpoint inhibitor therapy for cancer (PD-1 and PD-L1 blocking antibodies such as nivolumab, pembrolizumab, and avelumab) has been associated with new-onset type 1 diabetes and DKA.[48][49][23][50] It is estimated that up to 75% of people who develop immune checkpoint inhibitor-induced hyperglycaemia/diabetes present with DKA.[51]
Hypercortisolism leads to insulin resistance and may occasionally precipitate diabetic ketoacidosis in patients with concomitant diabetes; it more commonly precipitates hyperosmolar hyperglycaemic state.[52]
One US nationwide cohort study found that black patients with type 1 or type 2 diabetes had higher risks of hyperglycaemic crises compared with individuals in other racial and ethnic groups.[34] This disparity persisted after adjustment for key socioeconomic, clinical, and treatment-related factors, suggesting that additional intrinsic and extrinsic factors contribute to the increased susceptibility among black patients.[34]
Diabetic ketoacidosis (DKA) has also been increasingly documented as a presenting feature of newly diagnosed type 2 diabetes, a condition referred to as 'ketosis-prone diabetes'.[11][12] Epidemiological data suggest that people of African, Hispanic, and Asian origin are at greater risk.[11][13] Individuals with ketosis-prone type 2 diabetes often have obesity, a strong family history of type 2 diabetes, and evidence of insulin resistance.[11]
Among black patients with obesity who present with DKA, approximately 80% have type 2 diabetes, characterised by higher endogenous insulin secretion, absence of autoimmune markers, and a lack of HLA genetic association compared with lean patients with type 1 diabetes.[8]
Fasting can increase the risk of dehydration, hyperglycaemia, and ketoacidosis in people with diabetes, with the degree of risk varying according to diabetes type, treatment regimen, and other individual factors.[6] Adherence to a very low-carbohydrate or ketogenic diet while using a sodium-glucose cotransporter-2 (SGLT2) inhibitor or dual SGLT1/SGLT2 inhibitor may further increase the risk of diabetic ketoacidosis.[6][30]
Pregnancy is a physiologically ketogenic state, and individuals with pre-existing diabetes are at risk of developing diabetic ketoacidosis (DKA) at comparatively lower glucose levels, sometimes presenting as euglycaemic DKA.[6] Up to 2% of pregnancies complicated by pregestational diabetes (most commonly type 1 diabetes) are affected by DKA.[6] In contrast, the incidence of DKA in gestational diabetes is very low (<0.1%).[6]
Alcohol and substance use are recognised risk factors for diabetic ketoacidosis (DKA).[6] Excessive alcohol intake and use of illicit drugs may increase the risk of DKA associated with sodium-glucose cotransporter-2 (SGLT2) inhibitors and the dual SGLT1/SGLT2 inhibitor sotagliflozin.[30][6]
Cocaine use has been identified as an independent risk factor for recurrent DKA.[21] Cannabis use (and associated hyperemesis syndrome) has been associated with an increased risk of DKA in adults with type 1 diabetes.[6]
People with type 2 diabetes and comorbid dementia have a higher risk of developing diabetic ketoacidosis compared with those without dementia.[6]
The presence of mental health conditions - such as depression, bipolar disorder, and eating disorders - has been associated with an increased risk of hyperglycaemic crises.[6]
Use of this content is subject to our disclaimer