Sunteți pe pagina 1din 4

Review

Diabetic ketoacidosis in type 2 diabetes mellitus


A Puttanna1 Abstract
MBChB, MRCP, ST3, Diabetes and Endocrinology Diabetic ketoacidosis is usually associated with type 1 diabetes; however, it is increasingly
1 being recognised in type 2 diabetes. The three main mechanisms suggested are:
RNK Padinjakara insulinopaenia, elevation in counter-regulatory hormones as a stress response, and increase in
MBBS, MRCP, Consultant, Diabetes and Endocrinology
free fatty acids. This review aims to highlight the mechanism of diabetic ketoacidosis in type 2
1
Department of Diabetes and Endocrinology, Walsall
diabetes, the difference compared to its occurrence in type 1 diabetes, the main triggers and
Hospitals NHS Healthcare Trust, Walsall, UK its management.
The most common mechanism is relative insulin deficiency (insulinopaenia) and usual
triggers are non-concordance or infection. Treatment is exactly the same as in type 1 diabetes
Correspondence to: with intravenous fluid resuscitation and insulin, though the duration of treatment may not be
Amar Puttanna, MBChB, MRCP, ST3, Diabetes and as long. These patients are able to stop insulin following resolution of ketoacidosis and can be
Endocrinology, Department of Diabetes and managed on oral hypoglycaemic agents. It is important for clinicians to be aware of this
Endocrinology, Walsall Hospitals NHS Healthcare condition due to the increasing burden of type 2 diabetes and to avoid unnecessary treatment
Trust, Moat Road, Walsall, West Midlands WS2 9PS, with insulin in the long term. Copyright 2014 John Wiley & Sons.
UK; email: amarputtanna@doctors.org.uk
Practical Diabetes 2014; 31(4): 155158
Received: 4 November 2013
Accepted in revised form: 7 February 2014
Key words
DKA; ketoacidosis; type 2 diabetes; ketosis prone

Introduction acidaemia and raised blood glu-


Diabetic ketoacidosis (DKA) is a cose levels i.e. hyperglycaemia,4
well-known, life-threatening acute though hyperglycaemia may not
complication of type 1 diabetes. For always be present.5
a long time it has been considered It is a life-threatening condition
the hallmark of type 1 diabetes; how- with reported mortality rates in
ever, recently, its presence has been patients with type 1 diabetes of 2%
increasingly recognised in patients in the United Kingdom6 and 25%
with type 2 diabetes and a newer in the USA.7,8 However, prompt
entity called ketosis prone diabetes is recognition and management can
also commonly recognised. result in full recovery.
This paper reviews what is The incidence of DKA (regard-
currently known about DKA in less of underlying diabetes type and
patients with type 2 diabetes, and including first presentation) has
provides an insight into the possible been estimated as between 4.68
mechanisms and important practi- episodes per 1000 patients with
cal points for the clinician. diabetes in the USA,9 and there
There are documented instances is documented incidence of 11%
of ketosis and acidosis occurring of people with type 1 diabetes
in type 2 diabetes in the setting experiencing an episode of DKA
of excess alcohol intake, pregnancy in England.10
or anorexia;13 however, this The mechanism of DKA is due
paper looks at only confirmed to complete insulin deficiency, or
diabetic ketoacidosis in patients with relative insulin deficiency which is
type 2 diabetes. inadequate to prevent ketosis and
It is important to recognise that in the presence of excess counter-
DKA can occur in type 2 diabetic regulatory hormone.11
patients as this will have implica- The mechanism of the hypergly-
tions for acute management as well caemia in DKA is due to increased
as planning for further follow up glucose production from gluconeo-
and treatment, especially with the genesis and glycogenolysis, and
increasing prevalence of diabetes reduced uptake of glucose by
and the longer life expectancy of muscle and fat tissue.4,11
the population. Insulin is known to inhibit glu-
coneogenesis and glycogenolysis;
Diabetic ketoacidosis however, in insulin resistant states
Diabetic ketoacidosis is a condi- insulin is unable to effectively con-
tion characterised by ketonaemia, trol glucogenic enzymes thereby

PRACTICAL DIABETES VOL. 31 NO. 4 COPYRIGHT 2014 JOHN WILEY & SONS 155
Review
Diabetic ketoacidosis in type 2 diabetes mellitus

Authors DKA Type 2 DM: New Trigger Ethnicity BMI


cases no. (%) diagnosis: no. (kg/m2)

Newton, Raskin26 (USA) 138 30 (21.7) 5 Discontinue insulin/ White = 3 70% BMI >27
non-compliance 69.2% African = 17
Infection 48.4% Latino = 9
Other = 1

Wilson, et al.15 (USA) 17 17 (100) 0 Non-compliance 20% Apache Indians 24.94.4


Infection 47%
Hx of alcohol abuse in 94%

Balasubramanyam, 141 55 (39) 26 No trigger 50% Black = 28 77% BMI >25


et al.25 (USA) Hispanic = 20 51% BMI >30
White = 5
Asian = 2

Pitteloud, Philippe21 43 7 (16.3) Not clear No trigger 43% Caucasian Mean BMI 31.6
(Switzerland) Infection 28.5%
Pancreatitis 28.5%

Rao, et al.18 (India) 27 22 (81) Not clear Non-compliance 50% South Asian No data
Infection 33%
No trigger 14%

Jabbar, et al.17 (Pakistan) 57 57 (100) 8 Infection 63% South Asian No data

Chih-Hsun Chu, et al.16 137 98 (71.5) 24 Infection 48% South East Asian No data
(China) Non-compliance 19.4% (Chinese)

Table 1. Summary of key findings of studies analysing diabetic ketoacidosis (DKA) in type 2 diabetic patients1518,21,25,26

increasing glucose output from the discontinuation by professionals), studies have also looked at DKA in
liver.12 In insulin resistant states the initial presentation of diabetes Caucasian patients with type 2 dia-
body still remains sensitive to the and cardiovascular or cerebrovascu- betes.1921 This indicates that the
anti-lipolytic effects of insulin. lar events.4 development of DKA is not just
This is strengthened by data sug- From the above mechanisms it related to ethnic minorities as once
gesting that the amount of insulin could be assumed that it is possible was thought.
required to prevent lipolysis is for DKA to occur in patients The above studies were retro-
one-tenth of that required for glu- with type 2 diabetes when the spective ranging between one and
cose utilisation.13 insulin production is insufficient seven years in duration. All but
This is the reason why it had (or absent) to prevent ketone two16,21 classified type 2 diabetes
been thought that patients with production with or without precip- depending on whether the study
type 2 diabetes did not develop itating factors that is, relative population had been treated with
ketoacidosis. Type 2 diabetes is pre- insulin deficiency. diet or oral hypoglycaemic agents at
dominantly a disease of inadequate some point previously or, if new
insulin availability or increased DKA in type 2 diabetes onset, were controlled without
insulin resistance i.e. the bodys There have been many cases and insulin or were glutamic acid decar-
own insulin is insufficient for its studies looking at ketoacidosis boxylase (GAD)/islet cell antibody
needs. The residual beta-cell func- and its presence in patients with negative. The other two studies
tion in the pancreas of these type 2 diabetes. used basal C-peptide measurements
individuals could produce insulin There have been cases docu- to categorise patients as type 1 or
in sufficient amounts so as to pre- menting DKA in patients with type type 2 diabetes.
vent ketogenesis but inadequate for 2 diabetes as predominantly found The definitions of DKA were
the bodys glucose requirements, in ethnic minorities and specifi- based on variations of the American
thereby preventing build up of cally Afro-Caribbean populations Diabetes Association guidelines.
ketones in the blood stream. or indigenous populations of The parameters used were pH,
Various precipitating factors America.14,15 glucose levels, presence of ketones
have been acknowledged including However, studies have also (serum or urine), bicarbonate lev-
intercurrent infection/illness, omis- analysed DKA admissions in els and anion gap.
sion of regular insulin (either due Chinese,16 Pakistani17 and Indian18 Table 1 provides a review of the
to poor compliance or inadvertent populations, and further recent key findings of each of the studies.

156 PRACTICAL DIABETES VOL. 31 NO. 4 COPYRIGHT 2014 JOHN WILEY & SONS
Review
Diabetic ketoacidosis in type 2 diabetes mellitus

While predominantly found in Differences between DKA in type insulin.14,17,21 This may be related
adults with type 2 diabetes, it is 1 and type 2 diabetes to the recovery of beta-cell func-
interesting to note that DKA has Literature shows that DKA in tion once the acute hypergly-
also been described in children and patients with type 2 diabetes tends caemic episode has resolved.
adolescents with type 2 diabetes.22 to present with a less severe acidosis The importance of recognising
and patients are more likely to DKA as a feature of type 2 diabetes
Causes have normal potassium levels.25,26 lies in this finding. Understanding
The occurrence of DKA in type 2 Patients with type 2 diabetes and the nature of the diabetes type
diabetes has been thought to be DKA also tend to be older, have a ensures that patients are not unnec-
due to the presence of co-existing higher body mass index and a essarily continued on insulin. This
stressors, predominantly infections. shorter duration of diabetes with an can provide significant cost, eco-
Other reported causes include older age of onset.21 nomic and emotional benefit to
myocardial infarction, cerebrovas- Phenotypically, type 2 diabetes individuals due to the lifestyle
cular accidents, antipsychotic usage patients with DKA tend to have typ- restrictions and side effects that
and malignancy, such as pancreatic ical features of insulin resistance occur with insulin use.
adenocarcinoma.4,23,24 Sometimes such as large body habitus and Only one study has looked at
no stressor can be found and it can acanthosis nigricans (though this comparing inpatients with DKA
be the initial mode of presenta- is not present in all patients) who had either type 1 or type 2
tion.25,26 Another well documented compared to type 1 patients. They diabetes.32 The authors found that
cause is poor compliance with med- also have positive family history DKA was more likely to be associ-
ication either due to the patient and no autoimmune markers of ated with adverse outcomes in
themselves or inappropriate discon- diabetes, and may require larger patients with type 2 diabetes than
tinuation by medical profession- amounts of insulin to correct the in those with type 1 diabetes. They
als.15,16,18,26 The authors could hyperglycaemia.14,24 found a 30-day mortality rate of
not find any specific documented In keeping with type 1 diabetes, 11.9% in type 2 diabetes versus
reports of steroids triggering DKA the triggers for DKA development 2.4% in type 1 diabetes. This
in patients with pre-existing type 2 are similar, with infection and could possibly be related to the
diabetes; however, in one of the omission of insulin being the most presence of comorbidities and
studies mentioned above, one case common causes.1618,21 older age of presentation in the
of steroid induced DKA in the study Despite requiring insulin infu- type 2 group.
population was mentioned.17 sions to resolve ketoacidosis, many To date, no studies have
of these patients are able to stop looked at whether development
Mechanism in type 2 diabetes insulin therapy following resolution of DKA in patients with type 2 dia-
As described earlier, the occurrence of the DKA episode, and can be betes is associated with a poorer
of DKA in patients with type 1 dia- maintained on oral hypoglycaemic outcome in the long term. The
betes is due to the presence of agents or diet alone.14,17 risks of DKA remain the same in
insulinopaenia. A similar mechanism Improvement in C-peptide levels the acute phase regardless of dia-
has been thought to occur in long- following resolution of the DKA betes type.
standing type 2 diabetes patients due episode indicates improvement in
to complete loss of beta-cell function. beta-cell function;29,30 therefore Ketosis prone type 2 diabetes
However, this is not always the case as C-peptide measurement after Recently, the notion of ketosis
some patients present within a few glucagon administration may be of prone diabetes has been put for-
years of diagnosis where complete benefit in differentiating patients as ward as a way of classifying patients
beta-cell dysfunction is unlikely.26 being type 1 or type 2 when pre- with diabetes who are prone to
Studies suggest the cause could senting for the first time with DKA. ketone formation.
also be due to relative insulin This helps to identify whether there Study populations of various
deficiency arising from constant is any residual beta-cell function, ethnicities presenting with DKA
hyperglycaemia as a result of poor and can help predict whether or have been reviewed by Maldonado
control, and the presence of stres- not the patient requires further and Balasubramanyam et al. They
sors which cause increased lipolysis future insulin therapy.21,29,31 prospectively assessed patients
due to counter-regulatory hor- admitted with DKA with regard to
mones (glucagon, cortisol and Outcome of patients with DKA clinical and biochemical profiles.
growth hormone).24,27 It is also this and type 2 diabetes Their findings indicated that
hyperglycaemia which further The majority of patients who have patients presenting with DKA
blunts the bodys own insulin secre- an episode of DKA and have newly could be classified into four main
tion and reduces glucose removal diagnosed type 2 diabetes are groups depending on their charac-
by down-regulating glucose trans- able to discontinue insulin after teristics and presence of beta-cell
porter systems and even reducing the acute episode and remain on function or autoimmune antibod-
insulin gene transcription.28 The diet control or on oral hypogly- ies. They classified people into with
term glucose toxicity has often caemic agents. Studies show that or without autoimmune markers
been quoted when describing the between 5066.7% of patients (A+/A-) or having/not having
above mechanisms.14,28 admitted are able to remain off beta-cell function (b+/b-). Their

PRACTICAL DIABETES VOL. 31 NO. 4 COPYRIGHT 2014 JOHN WILEY & SONS 157
Review
Diabetic ketoacidosis in type 2 diabetes mellitus

15. Wilson C, et al. Ketoacidosis in Apache Indians


Key points with non-insulin-dependent diabetes mellitus.
Arch Intern Med 1997;157:2098100.
l Diabetic ketoacidosis is increasingly being reported in patients with type 2 diabetes and 16. Chih-Hsun Chu, et al. The occurrence of diabetic
ketoacidosis in type 2 diabetic adults. www.
can even be the initial mode of presentation in such individuals tsim.org.tw/journal/jour10-6/P10_230.PDF.
l Diagnostic criteria for diabetic ketoacidosis are the same irrespective of the type of [accessed 29 Aug 2013].
diabetes 17. Jabbar A, et al. Clinical characteristics and out-
comes of diabetic ketoacidosis in Pakistani adults
l There may not always be a trigger factor; however, infection and poor compliance have with type 2 diabetes mellitus. Diabet Med
been the most common reasons for developing diabetic ketoacidosis 2004;21:9203.
l Acidosis and hypokalaemia tend to be mild, but there is often a requirement for higher 18. Rao VD, et al. Clinical profile of diabetic ketoacido-
rates of insulin infusions in type 2 diabetes patients with diabetic ketoacidosis compared sis in adults. Health Renaissance 2012;10(2):806.
doi: http://dx.doi.org/10.3126/hren.v10i2.6569.
to patients with type 1 diabetes 19. Bagg W, et al. Diabetic ketoacidosis in adults at
l Intravenous insulin infusion is the mainstay of treatment and results in resolution of the Auckland Hospital, 19881996. Intern Med J 1998;
relative beta-cell functional deficit 28:6048.
20. Westphal SA. The occurrence of diabetic ketoaci-
l Most patients can be managed without insulin after the resolution of diabetic ketoacidosis dosis in non-insulin dependant diabetes and newly
and may remain off insulin for months to years diagnosed diabetic adults. Am J Med 1996;
101:1924.
21. Pitteloud N, Philippe J. Characteristics of Caucasian
classification had a sensitivity of however, its use into the medicine type 2 diabetic patients during ketoacidosis
99.4%, specificity of 95.9% and community as a whole is still to Schweiz Med Wochenschr 2000;130:57682.
22. Sapru A, et al. Prevalence and characteristics of
positive predictive value of 97.1% be accepted. type 2 diabetes mellitus in 918 year olds with
in predicting preserved beta-cell diabetic ketoacidosis. J Pediatr Endocrinol Metab
function in those patients admitted Declaration of interests 2005;18:86572.
23. Colli A, et al. Diabetic ketoacidosis associated
with DKA, and this could help There are no conflicts of interest with clozapine treatment. Diabetes Care 1999;22:
aid long-term management. They declared. Source of funding: none. 1767.
argued that old methods of classi- 24. Lin MV, et al. Diabetic ketoacidosis in type 2
fying diabetes are inadequate, and diabetics: A novel presentation of pancreatic
References adenocarcinoma. J Gen Intern Med 2010;25:
the new classification would help 1. Frise CJ, et al. Starvation ketoacidosis in preg- 36973.
divide patients with diabetes into nancy, Eur J Obstet Gynecol Reprod Biol 2013; 25. Balasubramanyam A, et al. New profiles of diabetic
those who are more likely to be at 167:17. ketoacidosis type 1 vs type 2 diabetes and the
2. Kamalakannan D, et al. Diabetic ketoacidosis in effect of ethnicity. Arch Intern Med
risk of ketosis, and help predict pregnancy. Postgrad Med J 2003;79:4547. 1999;159:231722.
long-term beta-cell function and 3. Fulop M. Alcoholic ketoacidosis. Endocrinol Metab 26. Newton CA, Raskin P. Diabetic ketoacidosis in type
requirement for insulin.30,33,34 Clin North Am 1993;22:20919. 1 and type 2 diabetes mellitus: clinical and bio-
4. Kitabchi AE, et al. Diabetic ketoacidosis and hyper- chemical differences. Arch Intern Med 2004;
The underlying mechanism in glycaemic hyperosmolar state. In International 164:192531.
ketosis prone diabetes is not textbook of diabetes mellitus, 3rd edn. De Fronzo 27. Poitout V. Glucolipotoxicity of the pancreatic cell:
known; however, increased suscep- RA, et al. (eds). 2004 John Wiley and Sons, 2004; myth or reality? Biochem Soc Trans 2008;
tibility to glucose desensitisation p1101. 36:9014.
5. Le Neveu F, et al. Euglycaemic ketoacidosis in 28. Robertson PR, et al. Differentiating glucose toxicity
has been suggested 35 whereby patients with and without diabetes. Pract Diabetes from glucose desensitization: a new message from
persistently high levels of glucose 2013;30:16771. the insulin gene. Diabetes 1994;43:10859.
in the blood stream lead to pan- 6. Wright J, et al. Diabetic ketoacidosis (DKA) in 29. Farber OK, Binder C. C-peptide response to
Birmingham, UK, 20002009: an evaluation of risk glucagon: a test for the residual beta-cell func-
creatic beta-cell desensitisation factors for recurrence and mortality. Br J Diabetes tion in diabetes mellitus. Diabetes 1977;
and lack of function at high glu- Vasc Dis 2009;9:27882. 26:60510.
cose levels. 7. Harris MI, Hamman RF. Diabetes in America. 30. Maldonado M, et al. Ketosis-prone diabetes:
Other studies have suggested the Washington DC: US Department of Health and Dissection of a heterogeneous syndrome using an
Human Sciences, 1985; ppXII-1-16. immunogenetic and beta cell functional classifica-
possibility of glucose-6-phosphate 8. Centers for Disease Control. Division of Diabetes tion, prospective analysis, and clinical outcomes.
dehydrogenase deficiency causing Translations: Diabetes Surveillance, 1991. J Clin Endocrinol Metab 2003;88:50908.
reduced beta-cell function in Washington DC: US Government Printing Office, 31. Misra S, et al. Diabetic ketoacidosis: not always
1992. due to type 1 diabetes. BMJ 2013;346:f3501.
patients with hyperglycaemia.36 9. Johnson DD, et al. Diabetic ketoacidosis in a 32. Barski L, et al. Comparison of diabetic ketoacidosis
A further study performed in community-based population. Mayo Clin Proc in patients with type-1 and type-2 diabetes melli-
2005 by Linfoot et al.37 looked at the 1980;55:838. tus. Am J Med Sci 2013;345:32630.
10. NHS Information Centre. National Diabetes Audit 33. Balasubramanyam A, et al. Accuracy and predictive
pathophysiology of DKA in ketosis 20082009. www.hscic.gov.uk/catalogue/PUB value of classification schemes for ketosis prone
prone type 2 diabetes mellitus by 02577/nati-diab-audi-08-09-exec-summ.pdf. diabetes. Diabetes Care 2006;29:25759.
obtaining blood samples from 11. American Diabetes Association. Hyperglycaemic 34. Kitabchi AE. Ketosis-prone diabetes A new sub-
patients presenting with DKA prior crises in patients with diabetes mellitus. Diabetes group of patients with atypical type 1 and type 2
Care 2002;25(Suppl 1):s100s108. diabetes? J Clin Endocrinol Metab 2003;88:
to intravenous insulin commence- 12. Barthel A, Schmoll D. Novel concepts in insulin reg- 50879.
ment, and suggested that the mech- ulation of hepatic gluconeogenesis. Am J Physiol 35. Umpierrez GE, et al. Ketosis-prone type 2 diabetes
anism was greater insulinopaenia Endocrinol Metab 2003;285:E68592. mellitus. Ann Intern Med 2006;144:3507.
13. Zierler KL, Rabinowitz D. Effect of very small con- 36. Sobngwi E, et al. High prevalence of glucose-6-
(i.e. relative insulin deficiency). centrations of insulin on forearm metabolism. phosphate dehydrogenase deficiency without
The notion of ketosis prone dia- Persistence of its action on potassium and free gene mutation suggests a novel genetic mecha-
betes and its classification is an fatty acids without its effect on glucose. J Clin nism predisposing to ketosis-prone diabetes. J Clin
interesting proposition and recent Invest 1964;43:95062. Endocrinol Metab 2005;90:444651.
14. Umpierrez G, et al. Diabetic ketoacidosis in 37. Linfoot P, et al. Pathophysiology of ketoacidosis
articles have highlighted this entity, obese African-Americans. Diabetes 1995;44: in type 2 diabetes mellitus. Diabet Med 2005;
bringing it into more prominence; 7905. 22:14149.

158 PRACTICAL DIABETES VOL. 31 NO. 4 COPYRIGHT 2014 JOHN WILEY & SONS

S-ar putea să vă placă și