Diabetic Ketoacidosis in Type 1 and Type 2
Diabetes Mellitus
Clinical and Biochemical Differences...
pared to the subgroup of patients with type 1 diabetes. Pa-
     1. Discontinue all subcutaneous insulin use.             ...
Table 1. Demographics for Subgroups of Patients Classified by Type and Previous Diagnosis of Diabetes

Table 4. Laboratory Findings in Patients Grouped According to Type of Diabetes
  and Whether Admission Was for Newly Diagn...
(39.8±13.3 hours, P = .01; and 36.0 ± 11.6 hours, P=.04;               dosis and a tendency for normal initial serum potas...
terregulatory hormones contribute to the hyperglyce-                   newly diagnosed group suggests that another process...
ferences in the degree of ketonuria between the sub-                   cal Center at Dallas, 5323 Harry Hines Blvd, G5.238...
Upcoming SlideShare
Loading in …5

Diabetic Ketoacidosis in Type 1 and Type 2 Diabetes Mellitus


Published on

Published in: Health & Medicine
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Diabetic Ketoacidosis in Type 1 and Type 2 Diabetes Mellitus

  1. 1. ORIGINAL INVESTIGATION Diabetic Ketoacidosis in Type 1 and Type 2 Diabetes Mellitus Clinical and Biochemical Differences Christopher A. Newton, MD; Philip Raskin, MD Background: Diabetic ketoacidosis (DKA), once thought tion use, 69.2% in the type 2 group. Infections were pres- to typify type 1 diabetes mellitus, has been reported to ent in 21.6% of the type 1 and 48.4% of the type 2 dia- affect individuals with type 2 diabetes mellitus. An analy- betes admissions. A total of 21% of patients with type 1 sis and overview of the different clinical and biochemi- diabetes and 70% with type 2 diabetes had a body mass cal characteristics of DKA that might be predicted be- index greater than 27. Although the type 1 diabetes group tween patients with type 1 and type 2 diabetes is needed. was more acidotic (arterial pH, 7.21±0.12 vs 7.27±0.08; P .001), type 2 diabetes patients required longer treat- Methods: We reviewed 176 admissions of patients with ment periods (36.0±11.6 vs 28.9±8.9 hours, P =.01) to moderate-to-severe DKA. Patients were classified as hav- achieve ketone-free urine. Complications from therapy ing type 1 or type 2 diabetes based on treatment history were uncommon. and/or autoantibody status. Groups were compared for differences in symptoms, precipitants, vital statistics, bio- Conclusions: A significant proportion of DKA occurs chemical profiles at presentation, and response to therapy. in patients with type 2 diabetes. The time-tested therapy for DKA of intravenous insulin with concomitant glu- Results: Of 138 patients admitted for moderate-to- cose as the plasma level decreases, sufficient fluid and severe DKA, 30 had type 2 diabetes. A greater propor- electrolyte replacement, and attention to associated prob- tion of the type 2 diabetes group was Latino American lems remains the standard of care, irrespective of the type or African American (P .001). Thirty-five admissions of diabetes. (19.9%) were for newly diagnosed diabetes. A total of 85% of all admissions involved discontinuation of medica- Arch Intern Med. 2004;164:1925-1931 A N EPISODE OF DIABETIC KE- present with DKA but have few other fea- toacidosis (DKA) was once tures associated with classic autoim- considered a hallmark fea- mune type 1 diabetes.9-12 ture that would differen- Recent epidemiologic studies13 esti- tiate individuals with type mate that hospitalizations for DKA have 1 diabetes mellitus from those with type increased during the past 2 decades. Part 2 diabetes mellitus. This is a concept that of this increased frequency of admissions has been perpetuated by such statements may be related to the increased preva- as “patients with NIDDM [non–insulin- lence of type 2 diabetes. With the dependent diabetes mellitus] are [not] ke- changes in the frequency of DKA and the tosis prone”1 and “many individuals with increased incidence of DKA in patients type 2 DM [diabetes mellitus] (previ- with type 2 diabetes mellitus, the ques- From The University Diabetes ously NIDDM) do not require insulin tion may be posed of whether there has Treatment Center, Parkland therapy to prevent ketoacidosis.”2 The been any change in the clinical or labora- Memorial Hospital, and clinical and biochemical characteristics of tory characteristics of the patients with Department of Internal DKA have been previously well de- DKA who present to the emergency de- Medicine, University of Texas scribed; however, most patients included partment. In this study, we present the Southwestern Medical Center at in these reports probably had type 1 dia- clinical and biochemical characteristics Dallas. Dr Newton is now with betes mellitus.3-5 More recently, there have of patients with type 2 diabetes com- the Department of Internal Medicine, The Brody School of been multiple reports of DKA occurring pared with type 1 diabetes admitted with Medicine at East Carolina in patients with type 2 diabetes melli- DKA to The University Diabetes Treat- University, Greenville, NC. The tus.6-8 Additionally, patients with idio- ment Center at Parkland Memorial Hos- authors have no relevant pathic type 1 diabetes, also referred to as pital, Dallas, Tex, as well as their re- financial interest in this article. type 1B, type 1.5, or Flatbush diabetes, will sponse to our treatment protocol. (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1925 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  2. 2. pared to the subgroup of patients with type 1 diabetes. Pa- 1. Discontinue all subcutaneous insulin use. tients were assigned to the type 2 diabetes group if they were 2. Measure blood glucose levels every hour. previously diagnosed as having diabetes and at some time in Measure urine ketones after each void or every 4 hours. their disease, other than a time consistent with the “honey- 3. Give D5%W intravenously using an infusion pump. moon period,” were managed with diet and exercise alone or 4. Make insulin solution using regular insulin to a concentration of 0.5 U/mL. with oral hypoglycemic agents or were noncompliant with their Give “piggyback” using an infusion pump into the D5%W infusion. insulin regimen for more than 3 weeks preceding admission. 5. Based on the hourly blood glucose determination, change each of the 2 Patients with newly diagnosed diabetes were assigned to the infusions according to the following schedule: type 2 group if they tested negative for autoimmune antibody or if they had phenotypic features of type 2 diabetes (such as Blood Glucose, Insulin Infusion D5%W, mg/dL mL/h obesity, acanthosis nigricans, or a family history of type 2 dia- U/h mL/h betes) and were noted to have a stressful event (eg, infection) ≤70∗ 0.5 1.0 250 that precipitated the episode of DKA. Based on recent evi- 71-100 1.0 2.0 225 dence that indicates similar biochemical profiles and physical 101-150 2.0 4.0 200 characteristics, patients with idiopathic type 1 diabetes (type 151-200 3.0 6.0 175 1B) were included in the type 2 diabetes subgroup for pur- 201-250 4.0 8.0 150 poses of analysis.15 Further analyses were performed compar- 251-300 6.0 12.0 100 ing patients diagnosed as having diabetes on admission for DKA 301-350 8.0 16.0 50 with those with a known history of diabetes. Additionally, pa- 351-400 10.0 20.0 0 tients with multiple admissions were compared with the group 401-450 12.0 24.0 0 of patients with only 1 admission during the year. 451-500 15.0 30.0 0 >500 20.0 40.0 0 ∗Give 20 mL of D50%W IVP and repeat CBG in 15 minutes. RESULTS A total of 138 patients accounted for the 176 admis- The insulin infusion algorithm used for the management of diabetic sions to Parkland Memorial Hospital’s University Dia- ketoacidosis. To convert blood glucose to millimoles per liter, multiply by betes Treatment Center for moderate-to-severe DKA. 0.0555. D5%W indicates dextrose 5% in water; D50%W, dextrose 50% in Eighty (58%) of the patients were male, 24 (17%) of the water; IVP, intravenous push; CBG, capillary blood glucose. patients were white, 65 (47%) were African American, 48 (35%) were Latino (mostly Mexican or Mexican Ameri- METHODS can), and 1 patient (1%) was Native American. The pa- tients were between the ages of 15 and 66 years (mean Of the 679 admissions to the Parkland Memorial Hospital Uni- age, 35.0±12.1 years). A total of 18.2% of the patients versity Diabetes Treatment Center between January and De- had a history of polysubstance abuse. A summary of the cember 2001, 176 admissions by 138 patients were for moderate- patient demographics for the subgroups analyzed is pre- to-severe DKA. The diagnosis of moderate-to-severe DKA was sented in Table 1. Differences between groups are noted made in the emergency department by the presence of 3 labo- for age and diabetes duration only. ratory findings: a plasma glucose level of 250 mg/dL or higher Symptoms documented to be present at the time of ( 13.9 mmol/L); a serum bicarbonate level of 15 mEq/L or lower, an arterial blood pH of 7.30 or lower, or a venous blood presentation are summarized in Table 2. The duration pH of 7.25 or lower; and moderate or large urinary ketones.14 of the symptoms, when they were reported to be pres- Patients were treated by residents of the University of Texas ent, is also shown. Although 42% of the patients admit- Southwestern Medical Center under the supervision of an en- ted to weight loss, only 78% of these patients were able docrinology fellow and an attending from the Division of En- to quantify the amount of loss, which was reported to docrinology. Intravenous insulin was administered according average 13.0 kg over 3 months. to the treatment algorithm shown in the Figure. This proto- A plausible explanation for the development of DKA col was begun as soon as possible after initial evaluation in the was identified in 90.3% of admissions. Thirty-six pa- emergency department. Additional hydration and electrolyte tients were newly diagnosed as having diabetes. Eighty- replacement were left to the discretion of the treating physi- five percent of the admissions by patients with a known cians, although following the American Diabetes Association practice guidelines was encouraged.14 The insulin infusion was history of diabetes followed a period of discontinuation discontinued 2 hours after the administration of subcutane- of medical therapies during the days preceding presen- ous insulin once patients had resolution of their metabolic sta- tation. Discontinuation of medication use, mostly insu- tus, including a ketone-free urine sample, and were able to tol- lin, was implicated in 90.2% of the admissions by the 16 erate oral feedings. patients with more than 1 admission during the year. Five These DKA admissions were reviewed retrospectively and of these patients had a history of polysubstance abuse. data collected on patient demographics, presenting symp- Infections were identified in 38% of all the patients and toms, precipitating causes of DKA, vital signs, biochemical pro- included urinary tract infections, upper respiratory tract files at presentation to the emergency department, amount of infections, pneumonia, cellulitis, and cutaneous ab- intravenous fluid administered, amount of insulin adminis- scesses. These infections were present in 22% of the pa- tered, duration of insulin infusion, time from presentation to resolution of urine ketones, insulin dose at discharge, and length tients with type 1 diabetes and 48% of patients with type of hospitalization. 2 diabetes. Pancreatitis was also present in 10% of the Comparisons between the groups were performed using admissions. Since patients with an acute myocardial in- the t test, Mann-Whitney rank sums tests, and the 2 test as farction were admitted to the cardiac care service rather appropriate. Unless otherwise indicated, all data are ex- than the University Diabetes Treatment Center, none of pressed as mean±SD. Patients with type 2 diabetes were com- the patients in this study had a myocardial infarction as (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1926 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  3. 3. Table 1. Demographics for Subgroups of Patients Classified by Type and Previous Diagnosis of Diabetes Previously Diagnosed Previously Diagnosed Newly Diagnosed Newly Diagnosed Variable Type 1 DM Type 2 DM Type 1 DM Type 2 DM No. of patients 78 25 30 5 Age, mean ± SD, y 32.5 ± 11.2 41.0 ± 12.5* 38.2 ± 13.3* 42.5 ± 3.3 Sex, No. (%) Male 44 (56) 15 (60) 18 (60) 3 (60) Female 34 (44) 10 (40) 12 (40) 2 (40) Ethnicity, No. (%) Non-Hispanic white 18 (23.1) 3 (12) 3 (10) 0 African American 36 (46.1) 12 (48) 12 (40) 5 (100) Latino American 24 (30.8) 9 (36) 15 (50) 0 Other 0 1 (4) 0 0 Duration of diabetes, mean ± SD, y 12.7 ± 10.0 5.6 ± 4.4† NA NA Abbreviations: DM, diabetes mellitus; NA, not applicable. *P .05 for comparison vs previously diagnosed type 1 DM group. †P .001 for comparison vs previously diagnosed type 1 DM group. Table 2. Frequency and Duration of Symptoms Reported Table 3. Frequency of Laboratory Data Findings by Patients Admitted With Moderate-to-Severe at the Time of Presentation Diabetic Ketoacidosis Frequency, % Frequency, Percentage Symptom Duration, Symptom of Admissions Mean ± SD, d Component Reference Range Low Normal High Polyuria 75.2 22.3 ± 33.6 Glucose 65-110 mg/dL 0 0 100 Polydipsia 74.4 23.1 ± 33.9 (3.6-6.1 mmol/L) Polyphagia 33.1 26.3 ± 34.2 Arterial pH 7.35-7.45 100.0 0 0 Weight loss 42.1 89.9 ± 97.7 Bicarbonate 22-31 mEq/L 100.0 0 0 Nausea 83.4 3.2 ± 3.1 Sodium 135-145 mEq/L 77.0 21.8 1.2 Vomiting 78.5 3.1 ± 3.1 Potassium 3.6-5.0 mEq/L 4.6 58.6 36.8 Abdominal pain 51.2 3.4 ± 3.3 Chloride 98-109 mEq/L 66.1 31.0 2.9 Blood urea 7-21 mg/dL 3.5 60.2 36.3 nitrogen (2.5-7.5 mmol/L) Creatinine 0.6-1.2 mg/dL 1.2 57.3 41.5 a precipitant for the development of DKA. Similarly, sur- (53.0-106.1 µmol/L) Calcium 8.4-10.2 mg/dL 12.4 68.2 19.4 gical stress was not an observed cause for the develop- (2.1-2.5 mmol/L) ment of DKA, since all of the patients in this study pre- Magnesium 1.4-1.8 mEq/L 10.0 67.5 22.5 sented to the emergency department presumably from (0.7-0.9 mmol/L) home. Phosphate 2.4-4.5 mg/dL 8.8 54.1 37.1 On presentation to the emergency department, pa- (0.8-1.5 mmol/L) tients were tachycardic, with an average heart rate of Amylase 29-108 U/L 23.6 59.9 16.5 Lipase 7-59 U/L 11.6 58.9 29.5 117 ± 19/min and had a slightly elevated systolic/ White blood 4.1-10.9 103/µL 0 44.3 55.7 diastolic blood pressure of 135 ± 23/85 ± 15 mm Hg. By cell count the time of discharge from the hospital, the patients had Neutrophils 45%-74% 0.6 35.1 64.3 normalization of the heart rate to 83 ± 9/min and nor- Lymphocytes 16%-45% 64.9 34.5 0.6 malization of systolic/diastolic blood pressure to 113±16/ Monocytes 4%-10% 22.8 69.0 8.2 73±10 mm Hg. Patients weighed 69.6 ± 18.3 kg (range, 43.1-147.2 kg; median, 65.3 kg) at the time of admis- sion to the hospital, having received an average of 3.7±1.5 L of intravenous fluid in the emergency department. Their Table 3 provides a summary of the distribution of weight increased by an additional 2.7±3.2 kg during the laboratory values for the patients at the time of presen- hospitalization. Patients newly diagnosed as having dia- tation with DKA. Patients had a hemoglobin A1c level of betes weighed more than those with a known history of 13.0%±2.5%. Patients with a clinical diagnosis of pan- diabetes (81.4± 22.7 kg vs 67.3 ± 16.5 kg, P = .02). Based creatitis at the time of admission accounted for 57% of on weight at the time of discharge from the hospital, 25% the patients with elevated amylase levels and 42% of the of the patients had a body mass index (a measure of weight patients with elevated lipase levels. Compared with those in kilograms divided by the square of height in meters) patients with normal or low potassium levels, the pa- greater than 27. Using their admission weights and re- tients with hyperkalemia at the time of presentation had ported amount of weight lost, 21% of patients with type correspondingly more severe acidosis. They also had 1 diabetes and 70% of patients with type 2 diabetes had higher plasma glucose levels, lower serum sodium lev- a body mass index greater than 27 before the onset of els, lower chloride levels, and higher calcium, magne- symptoms of uncontrolled diabetes. sium, and phosphate levels. Their blood urea nitrogen (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1927 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  4. 4. Table 4. Laboratory Findings in Patients Grouped According to Type of Diabetes and Whether Admission Was for Newly Diagnosed Diabetes Previously Diagnosed Previously Diagnosed Newly Diagnosed Newly Diagnosed Foster and Umpierrez Variable Type 1 DM Type 2 DM Type 1 DM Type 2 DM McGarry3 et al16 No. of 115 26 30 5 88 126 admissions Glucose, mg/dL 529.2 ± 212.4 (29.4 ± 11.8) 480.3 ± 159.8 (26.7 ± 8.9) 461.4 ± 192.3 (25.6 ± 10.7) 396.6 ± 191.1 (22.0 ± 10.6) 475 (26.4) 628 (34.9) (mmol/L) Arterial pH 7.21 ± 0.12 7.27 ± 0.08* 7.19 ± 0.14 7.28 ± 0.03† NA 7.16 Bicarbonate, 10.1 ± 5.1 10.2 ± 4.1 9.8 ± 5.0 12.0 ± 4.6 10 9 mEq/L Sodium, mEq/L 131.6 ± 5.0 129.5 ± 5.4 132.5 ± 5.8 130.6 ± 4.6 132 133 Potassium, 5.1 ± 0.9 4.5 ± 0.9* 4.4 ± 0.6* 4.8 ± 0.5 4.8 5.3 mEq/L Chloride, mEq/L 94.4 ± 8.1 93.1 ± 8.0 97.9 ± 6.3‡ 94.4 ± 6.9 NA 92 Blood urea 22.8 ± 13.3 (8.1 ± 4.7) 19.7 ± 12.5 (7.0 ± 4.5) 17.6 ± 12.3‡ (6.3 ± 4.4) 13.8 ± 4.8 (4.9 ± 1.7) 25 (8.9) 27 (9.6) nitrogen, mg/dL (mmol/L) Creatinine, 1.5 ± 0.9 (132.6 ± 79.6) 1.2 ± 0.7 (106.1 ± 61.9) 1.2 ± 0.5‡ (106.1 ± 44.2) 1.2 ± 0.3 (106.1 ± 26.5) NA NA mg/dL (µmol/L) Calcium, mg/dL 9.5 ± 0.9 (2.4 ± 0.2) 9.4 ± 0.9 (2.3 ± 0.2) 9.2 ± 1.0 (2.3 ± 0.2) 8.9 ± 0.7 (2.2 ± 0.2) NA NA (mmol/L) Magnesium, 1.7 ± 0.3 (0.9 ± 0.1) 1.7 ± 0.3 (0.8 ± 0.2) 1.7 ± 0.3 (0.8 ± 0.2) 1.5 ± 0.4 (0.7 ± 0.2) NA NA mEq/L (mmol/L) Phosphate, 5.0 ± 2.3 (1.6 ± 0.7) 4.0 ± 2.7 (1.3 ± 0.9) 3.9 ± 2.0‡ (1.3 ± 0.6) 2.9 ± 1.3 (0.9 ± 0.4) NA NA mg/dL (mmol/L) White blood 13.0 ± 6.3 13.3 ± 6.8 14.1 ± 7.3 10.0 ± 1.3†§ NA NA cells, 103/µL Neutrophils, % 77.6 ± 10.7 73.2 ± 10.8 74.0 ± 13.5 71.3 ± 12.9 NA NA Lymphocytes, % 14.4 ± 9.3 14.0 ± 10.1 14.9 ± 11.4 21.5 ± 10.8 NA NA Monocytes, % 6.0 ± 3.1 6.7 ± 2.9 6.7 ± 2.7 6.0 ± 3.4 NA NA Abbreviations: DM, diabetes mellitus; NA, not applicable. *P .01 vs previously diagnosed type 1 DM. †P .01 vs newly diagnosed type 1 DM. ‡P .05 vs previously diagnosed type 1 DM. §P .05 vs previously diagnosed type 2 DM. and creatinine levels were both elevated, but the ratio was which was lower in patients with a history of type 2 dia- similar between the 2 groups. betes than in those with type 1 diabetes (34.3±13.6 mg/dL Patients were classified as having type 2 diabetes, [0.9±0.4 mmol/L] vs 46.4±14.2 mg/dL [1.2±0.4 mmol/ either type 2 diabetes or idiopathic type 1 diabetes, in L], respectively, P .001) and in those who were newly 17.6% of the admissions. A greater percentage of these diagnosed as having classic type 1 diabetes (34.2±10.7 patients were of African American or Latino American mg/dL [0.9±0.3 mmol/L], P .001). No differences were ethnicity than the group of patients with type 1 diabetes observed in liver function test results. (P .001 by 2 test). Infections accounted for 15 of the The insulin-glucose infusion algorithm shown in the 31 admissions for the 30 patients in this group. Patients Figure was used in most admissions. Overall, patients had previously diagnosed as having type 2 diabetes were found cleared their urine ketones at an average of 31.7±11.1 to be noncompliant with their diabetes therapy in 69.2% hours after presenting to the emergency department and of the admissions. No significant differences were ob- 21.5±10.5 hours after being initiated on an insulin in- served in the frequency of symptom complaints in this fusion protocol. They received 9.5 ± 3.5 L of intrave- group compared with the type 1 diabetes group. This nous fluid during that time. Patients continued with the group of patients weighed more than the patients with intravenous insulin infusions for a total of 36.3 ± 28.4 type 1 diabetes (85.4±25.2 kg vs 66.1±14.4 kg, P=.002). hours and received a total of 200.3±130.1 U of insulin Other vital signs were not significantly different be- from this infusion and a total of 13.8±7.9 L of fluid. Com- tween the groups. pared with patients with a known history of type 1 dia- Laboratory data for the various subgroups ana- betes, who required 28.9±8.9 hours from the time of pre- lyzed are shown in Table 4. The laboratory data from 2 sentation to the emergency department and 152±47 U other reviews are also included in Table 4 for compari- of insulin to achieve a negative urine ketone status, pa- son. The cholesterol levels were similar in all groups ex- tients with newly diagnosed type 1 diabetes and those cept for the high-density lipoprotein cholesterol levels, with type 2 diabetes required both longer periods (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1928 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  5. 5. (39.8±13.3 hours, P = .01; and 36.0 ± 11.6 hours, P=.04; dosis and a tendency for normal initial serum potas- respectively) and more insulin (283 ± 156 U, P =.01; and sium levels, than patients with classic autoimmune type 175±69 U, P=.23; respectively) to achieve a negative urine 1 diabetes. ketone status. The patients with a known history of type It has been reported that up to 25% of cases of DKA 1 diabetes and those with type 2 diabetes achieved a nega- do not have an identified precipitating event.17,18 Fre- tive urine ketone status at similar times following initia- quently identified precipitating factors include intercur- tion of the intravenous insulin infusion protocol (19.1±8.4 rent illnesses such as infections or myocardial infarc- and 23.1±11.6 hours, respectively, P=.22). Patients newly tion, omission of or inadequate insulin therapy, and/or diagnosed as having type 1 diabetes required 31.7±13.3 newly diagnosed diabetes.17-19 Psychological issues and hours (P=.005) to achieve negative urine ketones after noncompliance are also known to be important factors, initiating the intravenous insulin protocol. especially in patients with repeated episodes of DKA as With regard to complications associated with treat- we have also found.17,19 Infections are reported to ac- ment of DKA, 4 patients had a blood glucose level of less count for nearly 30% to 40% of DKA and may be the most than 70 mg/dL ( 3.9 mmol/L) at some point during the frequent precipitant of DKA in white populations.16,18,19 insulin infusion. Three of these episodes occurred after Omission of insulin has been shown to be the most com- the patient received a subcutaneous injection of insulin mon precipitating factor in urban African American popu- in anticipation of discontinuing the intravenous insulin lations.16,20 We have found similar results in our popu- infusion but failed to tolerate the associated meal, attest- lation, which includes a large proportion of both African ing to the adequacy of the insulin-glucose infusion. Americans and Latino Americans, with more than 85% None of the episodes of hypoglycemia were reported to of the admissions being at least partially attributable to be clinically significant. Patients received an average of noncompliance with medical therapy. In our series, 32% 255 ± 213mM of potassium during their hospitaliza- of patients who had omitted their medications were found tions; however, since a specific protocol for potassium to also have an infection that often led the patient to dis- management was not provided to the treating physi- continue use of their insulin. Part of the explanation for cians, the frequency of patients developing hypokale- this behavior may be a fear of hypoglycemia if they are mia could not be appropriately assessed. Of note though, unable to eat well combined with a lack of understand- 8% of the patients did not receive any potassium supple- ing by the patient on sick day management of their dia- mentation. None of the patients developed cerebral edema, betes.21 Specific reasons for discontinuation of medical and there were no deaths. One patient developed a deep therapy among the patients in this study were not able venous thrombosis during his hospitalization. Another to be identified in the medical records. It is imperative patient was diagnosed as having adult respiratory dis- to remember that the finding of a plausible explanation tress syndrome within a few hours of presentation to the for the development of DKA, such as the omission of in- emergency department. sulin, does not preclude the existence of another pre- Patients remained hospitalized for an average of cipitating factor, such as infection. 4.5±3.3 days, consistent with the national average length The clinical picture of DKA typically involves poly- of stay for DKA of 4.5 days.13 Patients with known type uria, polydipsia, weight loss, weakness, air hunger with 1 diabetes were discharged home on an average daily in- Kussmaul respirations, vomiting, and abdominal pain. sulin dose of 1.02 ± 0.25 U/kg per day or 69 ± 22 U/d of Dehydration is almost always present in DKA but is less insulin. Patients with type 2 diabetes received 1.10±0.38 pronounced than in the hyperglycemic hyperosmolar syn- U/kg daily or 92 ± 39 U/d (P= .03 vs group with type 1 drome. The symptoms of polyuria, polydipsia, polypha- diabetes). Patients with newly diagnosed type 1 diabe- gia, and weight loss relate more to hyperglycemia than tes were discharged on a protocol of 89 ± 36 U/d or ketoacidosis. Although nonspecific, the symptoms of nau- 1.22±0.31 U/kg per day. Patients who used insulin be- sea, emesis, and abdominal pain are related to the pres- fore the admission for DKA were taking significantly less ence of ketonemia. It is clear from the duration of symp- insulin preceding the admission compared with their dis- toms listed in Table 2 that symptoms of osmotic diuresis charge dose (62 ± 23 U vs 72 ± 28 U, P = .002). (eg, polydipsia and polyuria) do not bring people to medi- cal attention, but nausea, vomiting, and abdominal pain COMMENT do. Unfortunately, the classic symptoms of DKA may not always be apparent.19 The percentages of symptoms Our results help illustrate the variability of clinical char- with which the patients presented in our study may be acteristics of patients who present with DKA and point an underestimate given that the data were collected ret- out the fact that patients with type 2 diabetes also de- rospectively and thus required that the health care team velop ketoacidosis. Even when restricting the popula- asked about the symptom and recorded it in the medi- tion to those patients with moderate or severe DKA, the cal record. patients have a variety of clinical and biochemical find- The laboratory data for our patients presented in ings as shown in Table 3. We found that some of the vari- Tables 3 and 4 are consistent with results that would be ability relates to the precipitating factors for DKA in that predicted for DKA. The biochemistry of DKA is in part patients with newly diagnosed diabetes have different bio- explained by the pathogenesis of this state. DKA occurs chemical findings compared with patients with a known in the setting of absolute or relative insulin deficiency history of diabetes as shown in Table 4. Additionally, we combined with increased levels of counterregulatory hor- found that patients with type 2 diabetes have a different mones, such as glucagon, catecholamines, cortisol, and biochemical presentation of DKA, with less severe aci- growth hormone. Insufficient insulin and elevated coun- (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1929 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  6. 6. terregulatory hormones contribute to the hyperglyce- newly diagnosed group suggests that another process, pos- mia through glycogenolysis, gluconeogenesis, and de- sibly related to the osmolarity, may be involved. creased peripheral use of glucose. Volume depletion is The newly diagnosed group did not have markedly probably the principal factor that results in the marked increased blood urea nitrogen levels and only very mildly hyperglycemia seen in DKA. The metabolic acidosis that increased creatinine levels, possibly explained by the develops in DKA is primarily the result of accumulation known interference of acetoacetate with the assay,18 which of the metabolic acids, acetoacetic acid and -hydroxy- might suggest that they are not as dehydrated as the pa- butyric acid, in the plasma. These ketoacids develop as tients with a known history of diabetes. The new diag- a consequence of the lipolysis and proteolysis that oc- nosis group required more than 4 L of additional fluid cur when elevated glucagon-insulin ratios cause in- and more than 12 additional hours of treatment with the creased hepatic levels of carnitine and decreased malonyl– insulin-glucose infusion to achieve a ketone-negative sta- coenzyme A concentrations.3 tus, as well as 10 L of additional fluid during the hospi- Other electrolyte abnormalities seem to develop as talization compared with the previously diagnosed group. a secondary effect of the hyperglycemia and metabolic However, the patients who also had pancreatitis ac- acidosis. The glycosuria that is present leads to at least counted for most of this increased fluid requirement. A mild dehydration and an associated increase in blood urea trend toward increased fluid requirements remained when nitrogen levels, as well as a mild increase in serum cre- the patients with pancreatitis were excluded from the atinine levels. Although serum sodium levels might be analysis (data not shown). Some of the normalcy of the expected to be increased in a dehydrated state, the os- laboratory results in the new diagnosis group could be motic effect of glucose draws water into the extracellu- related to fluid intake and maintenance of urine output lar space and tends to reduce the sodium concentration. before presentation to the emergency department; how- Potassium levels will tend to be high because of the physi- ever, this cannot be verified given the retrospective na- ologic compensation of the metabolic acidosis and the ture of our study. The status of the counterregulatory hor- hyperosmolarity and insulin deficiency present, al- mones may or may not be useful in understanding the though it is known that total body stores of potassium observed differences between these groups; however, since are depleted in DKA.19,22 It is therefore surprising to find these hormones are generally not of practical value in the that nearly 8% of the patients in this review did not re- clinical management of DKA, this study cannot address quire potassium supplementation during treatment of this possibility. their DKA. Although phosphate is severely depleted be- Patients with type 2 diabetes can develop more than cause of glycosuria, the plasma values will vary because mild ketosis and may experience moderate-to-severe DKA. of transmembrane shifts in the ion distribution caused It has been said that these patients require a precipitat- by metabolic acidosis and insulin deficiency. A stress- ing stressor; however, consistent with other reports, we induced leukemoid response is responsible for elevated found that 29% of the patients did not have a precipi- leukocyte counts. tant other than noncompliance with their medical A little more than one fourth of the patients had therapy.8 The tendency for the patients with type 2 dia- newly diagnosed diabetes, accounting for approxi- betes who develop DKA being Latino or African Ameri- mately 20% of the admissions. Consistent with other re- can compared with white may only be a reflection of the ports, newly diagnosed diabetes presenting with DKA was demographics of the overall patient population at Parkland diagnosed in a substantial number of overweight pa- Memorial Hospital. However, the possibility of a true eth- tients, a condition frequently associated with type 2 dia- nic variation in patients with type 2 diabetes who are prone betes.6,16 Our results suggest that patients with newly di- to develop DKA, possibly from a genetic predisposition, agnosed type 1 diabetes have a more normal electrolyte cannot be excluded. In contrast to other reports that com- profile than patients who have a known history of dia- pared DKA in type 1 and type 2 diabetes, we found a dif- betes despite similar degrees of acidosis. This is a bit sur- ference in the serum electrolyte concentrations and other prising, since they had similar glucose levels, pH levels, biochemical levels of these 2 groups.8 Some of the dif- apparent ketonuria, and duration of symptoms. It is ferences we observed may be secondary to differences in known that the severity of ketoacidosis is decreased when counterregulation and basal insulin levels in these groups. counterregulatory hormone levels are low.17 Further- Our patients with type 1 diabetes had had their disease more, it would be expected that patients with newly di- for a duration of 12.7 years, which is likely long enough agnosed diabetes would have a more intact counterregu- to place the patients at a time long after the “honey- latory hormone system than patients with a prolonged moon period” and at a time when they are likely to have duration of diabetes.23 As such, we might predict that the absolute insulin deficiency. The patients with type 2 dia- newly diagnosed diabetes would be associated with higher betes had a disease duration of 5.6 years and thus would glucagon and epinephrine levels, which would ulti- be expected to still have residual -cell activity.24 Al- mately induce a greater degree of ketogenesis and a more though the type 2 diabetes group had insufficient insu- severe picture of DKA. Hyperkalemia would be ex- lin to prevent DKA, it is theoretically possible that they pected in both groups of individuals, since the insulin had enough insulin to modulate the disease process. In deficiency is present in both, but would be predicted to support of this idea is the finding of lower potassium lev- be greater in the more acidotic group. Since the groups els in this group, which as noted herein are dependent had similar pH levels, the potassium and phosphate lev- on insulin presence or absence. The less severe acidosis els would be predicted to be similar if this were a major suggests a lower level of ketoacids; however, this was not contributing factor. Therefore, the normokalemia in the detected in this study. Although we did not observe dif- (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1930 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.
  7. 7. ferences in the degree of ketonuria between the sub- cal Center at Dallas, 5323 Harry Hines Blvd, G5.238, Dal- groups studied, we acknowledge that our method of mea- las, TX 75390-8858 (Philip.Raskin@UTSouthwestern.edu). suring ketoacids by semiquantitative urine testing fails to identify the ketoacid that is most prevalent in DKA REFERENCES and that a difference in -hydroxybutyric acid levels could exist, although this has not been studied. 1. Gregerman RI. Diabetes mellitus. In: Barker LR, Burton JR, Zieve PD, eds. Prin- The importance of recognizing DKA in a patient with ciples of Ambulatory Medicine. 5th ed. Philadelphia, Pa: Lippincott Williams & type 2 diabetes is illustrated by the time it took for these Wilkins; 1999:1023-1065. patients to achieve a negative urine ketone status. Once 2. Powers AC. Diabetes mellitus. In: Braunwald E, Fauci AS, Isselbacher KJ, et al, eds. Harrison’s Principles of Internal Medicine. 15th ed. New York, NY: McGraw- patients with either type 1 or type 2 diabetes were treated Hill Book Co; 2001:2109-2137. by the intravenous insulin infusion protocol, the dura- 3. Foster DW, McGarry JD. The metabolic derangements and treatment of diabetic tion of treatment required to achieve a negative urine ke- ketoacidosis. N Engl J Med. 1983;309:159-169. tone status was similar between the 2 groups. Patients 4. Atchley DW, Loeb RF, Richards DW Jr, Benedict EM, Driscoll ME. On diabetic acidosis: a detailed study of electrolyte balances following the withdrawal and with type 2 diabetes, however, required more time from reestablishment of insulin therapy. J Clin Invest. 1933;12:297-326. initial presentation to the emergency department to 5. Kreisberg RA. Diabetic ketoacidosis: new concepts and trends in pathogenesis achieve negative urine ketones. Some of this delay may and treatment. Ann Intern Med. 1978;88:681-695. be the result of a failure to suspect DKA in the “atypical 6. Umpierrez GE, Casals MMC, Gebhart SSP, Mixon PS, Clark WS, Phillips LS. Dia- patient.” betic ketoacidosis in obese African-Americans. Diabetes. 1995;44:790-795. 7. Valabhji J, Watson M, Cox J, Poulter C, Elwig C, Elkeles RS. Type 2 diabetes pre- Our results also show that the insulin-glucose in- senting as diabetic ketoacidosis in adolescence. Diabet Med. 2003;20:416-417. fusion algorithm outlined in the Figure can be used safely 8. Balasubramanyam A, Zern JW, Hyman DJ, Pavlik V. New profiles of diabetic ke- in the management of patients who present with DKA toacidosis: type 1 and type 2 diabetes and the effect of ethnicity. Arch Intern Med. irrespective of the type of diabetes. Complications that 1999;159:2317-2322. 9. Pinero-Pilona A, Litonjua P, Aviles-Santa L, Raskin P. Idiopathic type 1 diabetes ˜ ˜ ´ have been reported with management of DKA were in- in Dallas, Texas: a 5-year experience. Diabetes Care. 2001;24:1014-1018. frequent in our population in part because of the age of 10. Banerji MA, Chaiken RL, Huey H, et al. GAD antibody negative NIDDM in adult our patients but also because of the inclusion of a vari- black subjects with diabetic ketoacidosis and increased frequency of human leu- able glucose infusion algorithm in the treatment proto- kocyte antigen DR3 and DR4 Flatbush diabetes. Diabetes. 1994;43:741-745. col. Resolution of the acidosis caused by the presence of 11. Sobngwi E, Gautier JF. Adult-onset idiopathic type I or ketosis-prone type II dia- betes: evidence to revisit diabetes classification. Diabetologia. 2002;45:283- ketoacids is often used as a marker of resolution of DKA. 285. Reports that serum 3- -hydroxybutyric acid levels are 12. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Re- elevated in patients with type 2 diabetes who have nor- port of the expert committee on the diagnosis and classification of diabetes melli- mal fasting blood glucose levels raise concerns about their tus. Diabetes Care. 2003;26(suppl 1):S5-S20. 13. Centers for Disease Control, Division of Diabetes Translations. 1999 Diabetes utility in DKA management in patients with type 2 dia- Surveillance Report. Available at: http://www.cdc.gov/diabetes/statistics/survl99 betes.25,26 Our use of urine ketone testing does not pro- /Chap7/contents.htm. Accessed September 20, 2003. vide direct evidence of -hydroxybutyric acid levels, but 14. Kitabchi AE, Umpierrez GE, Murphy MB, et al. Management of hyperglycemic since acetoacetic acid levels remain elevated longer27 and crises in patients with diabetes. Diabetes Care. 2001;24:131-153. since urine levels may be positive even after resolution 15. Ramos-Roman M, Pinero-Pilona A, Aviles-Santa L, Raskin P. Pathophysiology ´ ˜ ˜ ´ of idiopathic type 1 diabetes. Diabetes. 2003;52(suppl):A337. of serum levels of acetoacetate, the resolution of keton- 16. Umpierrez GE, Kelly JP, Navarrete JE, Casals MMC, Kitabchi AE. Hyperglycemic uria would imply adequate resolution of ketonemia. crises in urban blacks. Arch Intern Med. 1997;157:669-675. Through our continuation of the insulin-glucose infu- 17. Kitabchi AE, Fisher JN, Murphy MB, Rumbak MJ. Diabetic ketoacidosis and the sion protocol until the urine was ketone free, we have hyperglycemic, hyperosmolar nonketotic state. In: Kahn CR, Weir GC, eds. Jos- lin’s Diabetes Mellitus. 13th ed. Philadelphia, Pa: Lea & Febiger; 1994:738-770. shown successful use of ketonuria as a marker of reso- 18. Jabbour SA, Miller JL. Uncontrolled diabetes mellitus. Clin Lab Med. 2001;21: lution of DKA in both type 1 and type 2 diabetes. Fi- 99-110. nally, given our results of the successful, almost unevent- 19. Ennis ED, Kreisberg RA. Diabetic ketoacidosis and the hyperglycemic hyperos- ful management of DKA, we have shown that in adults molar syndrome. In: LeRoith D, Taylor SI, Olefsky JM, eds. Diabetes Mellitus. the modern management of DKA with essentially no com- Philadelphia, Pa: Lippincott-Raven Publishers; 1996:276-286. 20. Musey VC, Lee JK, Crawford R, Klatka MA, McAdams D, Phillips LS. Diabetes in plications of treatment consists of effective intravenous urban African-Americans, I: cessation of insulin therapy is the major precipitat- insulin therapy with concomitant glucose administra- ing cause of diabetic ketoacidosis. Diabetes Care. 1995;18:483-489. tion when plasma glucose levels decrease below 300 21. American Diabetes Association. Standards of medical care for patients with dia- mg/dL (16.7 mmol/L), sufficient volume and electrolyte betes mellitus. Diabetes Care. 2003;26(suppl 1):S33-S50. 22. Adrogue HJ, Lederer ED, Suki WN, Eknoyan G. Determinants of plasma potas- ´ replacement, and careful attention to identification and sium levels in diabetic ketoacidosis. Medicine (Baltimore). 1986;65:163-172. management of associated problems, such as acute pan- 23. Bolli G, de Feo P, Compagnucci P, et al. Abnormal glucose counterregulation in creatitis and infection. insulin-dependent diabetes mellitus: interaction of anti-insulin antibodies and im- paired glucagon and epinephrine secretion. Diabetes. 1983;32:134-141. 24. Lebovitz HE. Insulin secretagogues: old and new. Diabetes Rev. 1999;7:139- Accepted for publication October 31, 2003. 153. We acknowledge the nursing and support staff of The 25. Okuda Y, Kawai K, Murayama Y, Yamashita K. Postprandial changes in plasma University Diabetes Treatment Center at Parkland Memo- ketone body and carnitine levels in normal and non-insulin-dependent diabetic rial Hospital for their dedicated commitment and care of the subjects. Endocrinol Jpn. 1987;34:415-422. 26. Laffel L. Ketone bodies: a review of physiology, pathophysiology and applica- patients with diabetes admitted to this service. tion of monitoring to diabetes. Diabetes Metab Res Rev. 1999;15:412-426. Correspondence: Philip Raskin, MD, Department of In- 27. Stephens JM, Sulway MJ, Watkins PJ. Relationship of blood acetoacetate and ternal Medicine, University of Texas Southwestern Medi- 3-hydroxybutyrate in diabetes. Diabetes. 1971;20:485-489. (REPRINTED) ARCH INTERN MED/ VOL 164, SEP 27, 2004 WWW.ARCHINTERNMED.COM 1931 Downloaded from www.archinternmed.com on October 21, 2010 ©2004 American Medical Association. All rights reserved.