Pseudonormoglycemia in Diabetic Ketoacidosis With Elevated Triglycerides MARK J. RUMBAK, MD,* THOMAS A. HUGHES, MD,t ABBAS E. KITABCHI, PhD, MD1_ A 24year-old newly dlagnosed male patient with diabetes presented with dlabetlc ketoacidosls (DKA) (pH 7.16, HC03 6.0) and extrame hyperlrlglycerldemla (239.36 mmol/L). The diagnosis of DKA was delayed because of the apparent depression of the true serum glucose (to 11 mmol/L). He was treated with Intravenous (IV) Insulin and rehydratlon, which normalized his pH, HC03, and trlglycerlde levels. To the authors’ knowledge, this Is both the highest triglyceride level recorded and the first report of a high triglyceride level as the apparent cause 01 a factitiously low glucose level. (Am J Emerg Med 1991:9:61-6X Copyright 0 1991 by W.6. Saunders Company)
Diabetic ketoacidosis (DKA) does not usually present any diagnostic problems.’ Poorly controlled diabetes mellitus may result in hypertriglyceridemia’ that can be seen in up to 50% of patients in DKA.3 It is well known that hypertriglyceridemia can cause pseudonormonatremia or pseudohyponatremia, and the ramifications of these effects have been recently reviewed.4 Similarly plasma concentration of glucose is decreased because of volume displacement by the high levels of circulating lipids.4 We are now reporting a case of DKA in which diagnosis was delayed because of pseudonormoglycemia, a high blood glucose measured in the near normal range. To our knowledge, this is the first case report of this nature. CASE REPORT A 24-year-old black man was admitted to the emergency department at the Regional Medical Center, Memphis (MED) in DKA. Two days before admission he had gone to the army recruting office; results of a urinalysis there showed that his urine contained 3+ glucose and 1 + ketones. The following day, his local medical doctor measured a blood glucose concentration of 7.5 mmol/L and the patient was sent home. When the patient presented to the MED, his blood glucose had risen to 10.82 mmol/L and his urine had 4+ glucose and 2+ ketones. Closer questioning disclosed a 2-month history of progressive weight loss of 10 kg, polyuria, polydypsia, polyphagia, and fatigue. His diet during this time had consisted of eggs and cheese, sausages, bacon and eggs, porkchops, and pigs’ feet. There was no history of cardiovascular disease or hypertension, and he had no symptoms of renal, hepatic, pancreatic, peptic ulcer, or thyroid disease. There was no previous surgery or cancer history. He was a muscular man who did not drink alcohol or smoke ciga-
rettes. There were no symptoms of diabetic peripheral neuropathy, retinopathy or autonomic neuropathy. The patient’s father and material grandmother had type II diabetes, controlled by diet. The patient was not taking any medication.
PHYSICAL EXAMINATION The patient was a well developed mildly dehydrated black male in mild respiratory distress. His vital signs were pulse rate, 120 beats/ min; blood pressure, 110/80 mm Hg right arm lying, !90/80mm Hg right arm standing; temperature 37.2”C; respiratory rate, 36 breaths/ min, Kussmaul’s in character. His weight was 54.5 kg and height was 1.5 m. He was alert and orientated to person, place and time. The head and neck examination showed dry mucus membranes and a slight increase in tissue turgor and diminished sweating. He had no arcus comealis, thickened tendons or eruptive xanthomata. Pupils were equally reactive to light and accommodation and examination of his fundus showed lipemia retinalis. His sclera were anicteric. Chest, heart, abdomen were normal, and secondary sexual characteristics were normal male. There was no evidence of peripheral or autonomic neuropathy, and his pulses were present without bruits. Initial laboratory results and those following treatment are shown in Table 1. The laboratory had reported extremely cloudy serum. The MED did not centrifuge the serum to attempt to clear some triglyceride. Serum amylase and lipase were not raised.
HOSPITAL COURSE After the diagnosis of DKA was suspected, the patient received 1 L of normal saline in the first hour. By the end of this infusion, laboratory results were back and the diagnosis was made. The patient was transferred to the intensive care unit and haIf normalsaline was given at a rate of 150 mL/h. Intravenous (IV) insulin was administered as follows: 14 units regular IV bolus and 14 units subcutaneous (SC) followed by 7 Uih continuous IV infusion. No bicarbonate or phosphate was given. The endocrine service was aware of possible pseudonormoglycemia and recommended decreasing the apparent blood glucose to between 4.5 mmol/L and 6 mmol/L (instead of the usual 8.5 to 11 mmol!L). This was obtained by adding 5% dextrose to the half normal saline. Once the patient was out of DKA as defined by blood glucose less 8.35 mmol/L, pH > 7.30 and bicarbonate > 15 mg/L, he was changed to long-acting human insulin twice a day. He was able to eat and was given a 2,200 calorie American Diabetes Association low-fat diet and a sliding scale of regular human insulin before meals. All blood sugars ranged between 5.5 and 8.5 mmoYL while in the hospital. The patient subsequently refused all other medication.
DISCUSSION From the *Department of Medicine, Division of Pulmonary and Critical Care. and the tDeoartment of Medicine, Division of Endocrinology gnd Metadoliim and Lipid Clinic, University of Tennessee, Memphis, TN. Manuscript received April 9, 1990; revision accepted August 1, 1990. Address reprint requests to Dr Rumbak: 956 Court Avenue, Room H314, Memphis, TN 36163. Copyright 0 1991 by W.B. Saunders Company 0735-6757/91/0901-0016$5.00/0
Lipoprotein lipase (LPL) is the major enzyme responsible for removal of triglycerides from plasma. The function of this enzyme has been extensively reviewed.* The congenital forms of LPL deficiency have been well characterized’ and present as a type I hyperlipidemia. Hyperlipidemic phenotypes may also be seen if the activator of lipoprotein lipase, (lipoprotein C-II) is deficient, if there is a circulating inhibitor of lipoprotein lipase (II), or if autoimmune hyperchylo61
l
Invalid.
Urine Glucose (mmol/L) Ketones (kmol/L)
Arterial Blood Gas PH PCO, Calculated bicarb Percentage saturation PO*
Serum Triglyceride (mmol/L) Glucose (mmol/L) Sodium Potassium Chloride Measured CO2 BUN (mmol/L) Creatinine (ikmol/L) Anion gap Acetoacetic acid (@mol/L) Lactate
Date Time
22 11754 Neg
f
i.9
239.35 77 175 4.3 86
2/24 1334
*
>55.5 >7836
7.76 13 6 98 147
19 77754
0.44 179 3.8 95 5 3.6
1730
79
7.1 120 3.3 96 5 3.9 f
2005
7.23 19 0 98 121
14
7.33 718 3.5 97 7 3.2 I
2145
7.24 29 72 97
17637
* 9
57.13 6.9 779 3.4 100 10 3.2
2!25 0230
9
5.44 120 3.2 700 11 2.8 f
0430
>74 >7836
7.25 33 14 98 714
9
4.67 121 3.4 707 11 2.5 f
0600
7
6.27 126 3.5 106 13 3.6 f
0600
7
5.0 127 3.6 707 13 2.5 t
1000
TABLE1. FlowSheet Showing the Patient’s Major Biochemical Indices on Admission and Their Responses to Treatment
4
4.83 726 3.6 109 13 3.2 f
7200
5 4890
*
4.4 726 3.5 107 14 2.5
7400
.
5 Neg
17.40 8 129 3.9 106 18 2.8
2/26 0435
7.6 134 3.6 106 23 3.6 97 5
2/27 0320
9.51 6.7 135 4.1 105 25 3.6 67.9 5
2128 0525
7.5 136 4.8 97 37 2.8 97 8
3l7 0750
5.7 137 4.0 98 33 -3.2 79.5 6
3/2 0420
7.0 138 4.6 104 27 -2.8 97 7
3124 0700
w
RUMBAK, HUGHES, AND KITABCHI m PSEUDONORMOGLYCEMIA IN DKA
micron exists.’ In the genetically increased very-low-density lipoprotein (VLDL) states, the LPL3 may be depressed or normal.’ The acquired forms of lipoprotein lipase deficiency include uncontrolled diabetes mellitus, severe hypothyroidism, and chronic renal failure or nephrotic syndrome.’ LPL is insulin-dependent. Therefore, in extreme cases of insulin deficiency3 there is a very low level of lipase activity and the excessive dietary intake of triglycerides cannot be adequately cleared from the plasma, resulting in hypertriglyceridemia. The severity of the insulin deficiency seems to correlate inversely with the lipoprotein lipase activitye5 In our patient, one must assume the large fat intake and insulin deficiency aggravated the hypertriglyceridemia. (He was a virgin diabetic). The overproduction of VLDL and decreased clearance of glycosylased LDL and VLDL could have accentuated this hypertriglyceridemia. The extreme hypertriglyceridemia in our patient resulted in factitiously lower serum sodium, glucose, pH, anion gap and bicarbonate. Comparing the measured bicarbonate with the calculated bicarbonate in the arterial blood gas (Table l), the former was consistently lower. Whether this is significant or not requires further study. The diagnosis of ketoacidosis was delayed because of a near normal serum glucose. Hilton et al6 have suggested measuring the urinary anion gap (which may be increased) when the serum gap is normal or near normal in a patient with severe hyperlipidemia, suspected of having DKA. Morris et al’ have shown that blood glucose and urine glucose do not always correlate. However, in this case, the urine glucose rather than blood glucose may be a better index of response to intravenous insulin and should be monitored more closely than usual. Management of hypertriglyceridemia consists of treating the DKA by the usual methods.’ Intravenous insulin was administered as described above,’ except that the serum glucose was clamped betwen 4.5 and 6 mmol/L. The high levels of triglycerides decreased by >90% in 2 days and were nearly normal when the patient returned for follow-up 2 weeks later. We were surprised at this, as we expected a mixed cause of this massive number of triglyceridemia. His fasting blood glucose at that time was 6 mmol/L. Some have suggested using plasma exchange’ as an initial
63
treatment, but our patient and the patient reported by Blackett et al9 were treated adequately with just routine intravenous insulin therapy.’ If hyperlipidemia persists for several months after DKA has resolved, then diet and pharmacologic therapy will probably need to be considered, as it is likely that in these cases the cause is underlying genetic form of hyperlipidemia.5 In summary, we present a patient whose diagnosis of DKA was delayed because of extremely high serum triglycerides causing pseudonormoglycemia. Clues to the correct diagnosis include lipemia retinalis, lipemic serum, and a urine anion gap significantly larger than the normal serum anion gap. The authors thank Vicky Franke for typing the manuscript.
REFERENCES 1. Kitabchi AE, Murphy MB: Diabetic ketoacidosis and hyperosmolar hyperglycemic nonketotic coma. Med Clin N Amer 1988;72:1545-1563 2. Granda OP: Pathogenesis of macrovascular disease including the influence of lipids. In Marble A, Krall LP, Bradley RF, Christlfeb AR, Soelder 58(eds). Joslin’s Diabetes Mellitus~12th ed. Philadeohia. PA, Lea & Cebiaer. 1985. D 224 3. Bagdaie JD, Porte, BIerman-EL: Diabetic lipemia: a form of acquired fat induced lipemia. N Engl J Med 1967;276:427-433 4. Weisberg LS. Pseudohyponatremia: A reappraisal. Am J Med 1989;896:315-318 5. Eckel RH: Lipoprotein lipase: A multifunctional enzyme relevant to common metabolic diseases. N Engl J Med 1989; 320:1060-1068 6. Hilton JG, Vandenbroucke AC, Josse RG, Buckley GC, Halperin ML: The urine anion gap: The critical clue to resolve a diagnostic dilemma in a patient with ketoacidosis. Diab Care 1984;27:286-490 7. Morris CR, McGee JA, Kitabchi AE: Correlation between plasma and urine glycose in diabetes. Ann Int Med 1981;94:469471 8. Gerard A, Schooneman F, Guine JM, Roche G, Canton P, Dureux JB, Janot C, Streiff F: Treatment by plasma exchange of a patient with hyperlepidemia and diabetic ketoacidosis with lesional pulmonary edema and acute pancreatitis. VOX Sang 1982:43:147-150 9. Blackett PR, Holcombe JH, Alaupovic P, llsmire JD: Plasma lipids and apolipoprotein in a 13 year old boy with diabetic ketoacidosis and extreme hyperlipidemia. Am J Med SC 1986; 2911342-346
Diabetic ketoacidosis (DKA) does not usually present any diagnostic problems.’ Poorly controlled diabetes mellitus may result in hypertriglyceridemia’ that can be seen in up to 50% of patients in DKA.3 It is well known that hypertriglyceridemia can cause pseudonormonatremia or pseudohyponatremia, and the ramifications of these effects have been recently reviewed.4 Similarly plasma concentration of glucose is decreased because of volume displacement by the high levels of circulating lipids.4 We are now reporting a case of DKA in which diagnosis was delayed because of pseudonormoglycemia, a high blood glucose measured in the near normal range. To our knowledge, this is the first case report of this nature. CASE REPORT A 24-year-old black man was admitted to the emergency department at the Regional Medical Center, Memphis (MED) in DKA. Two days before admission he had gone to the army recruting office; results of a urinalysis there showed that his urine contained 3+ glucose and 1 + ketones. The following day, his local medical doctor measured a blood glucose concentration of 7.5 mmol/L and the patient was sent home. When the patient presented to the MED, his blood glucose had risen to 10.82 mmol/L and his urine had 4+ glucose and 2+ ketones. Closer questioning disclosed a 2-month history of progressive weight loss of 10 kg, polyuria, polydypsia, polyphagia, and fatigue. His diet during this time had consisted of eggs and cheese, sausages, bacon and eggs, porkchops, and pigs’ feet. There was no history of cardiovascular disease or hypertension, and he had no symptoms of renal, hepatic, pancreatic, peptic ulcer, or thyroid disease. There was no previous surgery or cancer history. He was a muscular man who did not drink alcohol or smoke ciga-
rettes. There were no symptoms of diabetic peripheral neuropathy, retinopathy or autonomic neuropathy. The patient’s father and material grandmother had type II diabetes, controlled by diet. The patient was not taking any medication.
PHYSICAL EXAMINATION The patient was a well developed mildly dehydrated black male in mild respiratory distress. His vital signs were pulse rate, 120 beats/ min; blood pressure, 110/80 mm Hg right arm lying, !90/80mm Hg right arm standing; temperature 37.2”C; respiratory rate, 36 breaths/ min, Kussmaul’s in character. His weight was 54.5 kg and height was 1.5 m. He was alert and orientated to person, place and time. The head and neck examination showed dry mucus membranes and a slight increase in tissue turgor and diminished sweating. He had no arcus comealis, thickened tendons or eruptive xanthomata. Pupils were equally reactive to light and accommodation and examination of his fundus showed lipemia retinalis. His sclera were anicteric. Chest, heart, abdomen were normal, and secondary sexual characteristics were normal male. There was no evidence of peripheral or autonomic neuropathy, and his pulses were present without bruits. Initial laboratory results and those following treatment are shown in Table 1. The laboratory had reported extremely cloudy serum. The MED did not centrifuge the serum to attempt to clear some triglyceride. Serum amylase and lipase were not raised.
HOSPITAL COURSE After the diagnosis of DKA was suspected, the patient received 1 L of normal saline in the first hour. By the end of this infusion, laboratory results were back and the diagnosis was made. The patient was transferred to the intensive care unit and haIf normalsaline was given at a rate of 150 mL/h. Intravenous (IV) insulin was administered as follows: 14 units regular IV bolus and 14 units subcutaneous (SC) followed by 7 Uih continuous IV infusion. No bicarbonate or phosphate was given. The endocrine service was aware of possible pseudonormoglycemia and recommended decreasing the apparent blood glucose to between 4.5 mmol/L and 6 mmol/L (instead of the usual 8.5 to 11 mmol!L). This was obtained by adding 5% dextrose to the half normal saline. Once the patient was out of DKA as defined by blood glucose less 8.35 mmol/L, pH > 7.30 and bicarbonate > 15 mg/L, he was changed to long-acting human insulin twice a day. He was able to eat and was given a 2,200 calorie American Diabetes Association low-fat diet and a sliding scale of regular human insulin before meals. All blood sugars ranged between 5.5 and 8.5 mmoYL while in the hospital. The patient subsequently refused all other medication.
DISCUSSION From the *Department of Medicine, Division of Pulmonary and Critical Care. and the tDeoartment of Medicine, Division of Endocrinology gnd Metadoliim and Lipid Clinic, University of Tennessee, Memphis, TN. Manuscript received April 9, 1990; revision accepted August 1, 1990. Address reprint requests to Dr Rumbak: 956 Court Avenue, Room H314, Memphis, TN 36163. Copyright 0 1991 by W.B. Saunders Company 0735-6757/91/0901-0016$5.00/0
Lipoprotein lipase (LPL) is the major enzyme responsible for removal of triglycerides from plasma. The function of this enzyme has been extensively reviewed.* The congenital forms of LPL deficiency have been well characterized’ and present as a type I hyperlipidemia. Hyperlipidemic phenotypes may also be seen if the activator of lipoprotein lipase, (lipoprotein C-II) is deficient, if there is a circulating inhibitor of lipoprotein lipase (II), or if autoimmune hyperchylo61
l
Invalid.
Urine Glucose (mmol/L) Ketones (kmol/L)
Arterial Blood Gas PH PCO, Calculated bicarb Percentage saturation PO*
Serum Triglyceride (mmol/L) Glucose (mmol/L) Sodium Potassium Chloride Measured CO2 BUN (mmol/L) Creatinine (ikmol/L) Anion gap Acetoacetic acid (@mol/L) Lactate
Date Time
22 11754 Neg
f
i.9
239.35 77 175 4.3 86
2/24 1334
*
>55.5 >7836
7.76 13 6 98 147
19 77754
0.44 179 3.8 95 5 3.6
1730
79
7.1 120 3.3 96 5 3.9 f
2005
7.23 19 0 98 121
14
7.33 718 3.5 97 7 3.2 I
2145
7.24 29 72 97
17637
* 9
57.13 6.9 779 3.4 100 10 3.2
2!25 0230
9
5.44 120 3.2 700 11 2.8 f
0430
>74 >7836
7.25 33 14 98 714
9
4.67 121 3.4 707 11 2.5 f
0600
7
6.27 126 3.5 106 13 3.6 f
0600
7
5.0 127 3.6 707 13 2.5 t
1000
TABLE1. FlowSheet Showing the Patient’s Major Biochemical Indices on Admission and Their Responses to Treatment
4
4.83 726 3.6 109 13 3.2 f
7200
5 4890
*
4.4 726 3.5 107 14 2.5
7400
.
5 Neg
17.40 8 129 3.9 106 18 2.8
2/26 0435
7.6 134 3.6 106 23 3.6 97 5
2/27 0320
9.51 6.7 135 4.1 105 25 3.6 67.9 5
2128 0525
7.5 136 4.8 97 37 2.8 97 8
3l7 0750
5.7 137 4.0 98 33 -3.2 79.5 6
3/2 0420
7.0 138 4.6 104 27 -2.8 97 7
3124 0700
w
RUMBAK, HUGHES, AND KITABCHI m PSEUDONORMOGLYCEMIA IN DKA
micron exists.’ In the genetically increased very-low-density lipoprotein (VLDL) states, the LPL3 may be depressed or normal.’ The acquired forms of lipoprotein lipase deficiency include uncontrolled diabetes mellitus, severe hypothyroidism, and chronic renal failure or nephrotic syndrome.’ LPL is insulin-dependent. Therefore, in extreme cases of insulin deficiency3 there is a very low level of lipase activity and the excessive dietary intake of triglycerides cannot be adequately cleared from the plasma, resulting in hypertriglyceridemia. The severity of the insulin deficiency seems to correlate inversely with the lipoprotein lipase activitye5 In our patient, one must assume the large fat intake and insulin deficiency aggravated the hypertriglyceridemia. (He was a virgin diabetic). The overproduction of VLDL and decreased clearance of glycosylased LDL and VLDL could have accentuated this hypertriglyceridemia. The extreme hypertriglyceridemia in our patient resulted in factitiously lower serum sodium, glucose, pH, anion gap and bicarbonate. Comparing the measured bicarbonate with the calculated bicarbonate in the arterial blood gas (Table l), the former was consistently lower. Whether this is significant or not requires further study. The diagnosis of ketoacidosis was delayed because of a near normal serum glucose. Hilton et al6 have suggested measuring the urinary anion gap (which may be increased) when the serum gap is normal or near normal in a patient with severe hyperlipidemia, suspected of having DKA. Morris et al’ have shown that blood glucose and urine glucose do not always correlate. However, in this case, the urine glucose rather than blood glucose may be a better index of response to intravenous insulin and should be monitored more closely than usual. Management of hypertriglyceridemia consists of treating the DKA by the usual methods.’ Intravenous insulin was administered as described above,’ except that the serum glucose was clamped betwen 4.5 and 6 mmol/L. The high levels of triglycerides decreased by >90% in 2 days and were nearly normal when the patient returned for follow-up 2 weeks later. We were surprised at this, as we expected a mixed cause of this massive number of triglyceridemia. His fasting blood glucose at that time was 6 mmol/L. Some have suggested using plasma exchange’ as an initial
63
treatment, but our patient and the patient reported by Blackett et al9 were treated adequately with just routine intravenous insulin therapy.’ If hyperlipidemia persists for several months after DKA has resolved, then diet and pharmacologic therapy will probably need to be considered, as it is likely that in these cases the cause is underlying genetic form of hyperlipidemia.5 In summary, we present a patient whose diagnosis of DKA was delayed because of extremely high serum triglycerides causing pseudonormoglycemia. Clues to the correct diagnosis include lipemia retinalis, lipemic serum, and a urine anion gap significantly larger than the normal serum anion gap. The authors thank Vicky Franke for typing the manuscript.
REFERENCES 1. Kitabchi AE, Murphy MB: Diabetic ketoacidosis and hyperosmolar hyperglycemic nonketotic coma. Med Clin N Amer 1988;72:1545-1563 2. Granda OP: Pathogenesis of macrovascular disease including the influence of lipids. In Marble A, Krall LP, Bradley RF, Christlfeb AR, Soelder 58(eds). Joslin’s Diabetes Mellitus~12th ed. Philadeohia. PA, Lea & Cebiaer. 1985. D 224 3. Bagdaie JD, Porte, BIerman-EL: Diabetic lipemia: a form of acquired fat induced lipemia. N Engl J Med 1967;276:427-433 4. Weisberg LS. Pseudohyponatremia: A reappraisal. Am J Med 1989;896:315-318 5. Eckel RH: Lipoprotein lipase: A multifunctional enzyme relevant to common metabolic diseases. N Engl J Med 1989; 320:1060-1068 6. Hilton JG, Vandenbroucke AC, Josse RG, Buckley GC, Halperin ML: The urine anion gap: The critical clue to resolve a diagnostic dilemma in a patient with ketoacidosis. Diab Care 1984;27:286-490 7. Morris CR, McGee JA, Kitabchi AE: Correlation between plasma and urine glycose in diabetes. Ann Int Med 1981;94:469471 8. Gerard A, Schooneman F, Guine JM, Roche G, Canton P, Dureux JB, Janot C, Streiff F: Treatment by plasma exchange of a patient with hyperlepidemia and diabetic ketoacidosis with lesional pulmonary edema and acute pancreatitis. VOX Sang 1982:43:147-150 9. Blackett PR, Holcombe JH, Alaupovic P, llsmire JD: Plasma lipids and apolipoprotein in a 13 year old boy with diabetic ketoacidosis and extreme hyperlipidemia. Am J Med SC 1986; 2911342-346
