Winston, David C. M.D., Ph.D.
Although not numerous, cases of nonfatal accidental and intentional insulin overdose in both diabetic and nondiabetic patients have been reported (1-6). In the English literature, fatal self-inflicted insulin overdose in nondiabetic individuals is even less common, with six cases described in detail (6-11) and two others mentioned but not completely described (6). Critchley et al. described 15 cases of insulin self-injection in both diabetic and nondiabetic patients with two deaths; however, they did not state the diabetes status of the two decedents (2). Without an early suspicion, these cases may be missed because of incomplete scene investigation or improper collection and storage of postmortem specimens. We report four additional cases of suicide by insulin injection in decedents without a history of insulin-dependent diabetes mellitus. Also discussed are recommendations for the appropriate collection and storage of postmortem specimens for detection of insulin and C peptide.
MATERIALS AND METHODS
The case files of the Office of the Medical Investigator (Albuquerque, NM) were searched from 1975 to 1997. Insulin was used as a keyword, and all nonnatural deaths were searched. An initial search identified 26 cases; however, only four involved nondiabetic patients. Case histories are listed below.
Insulin and C Peptide Analysis
When performed, insulin and C peptide levels were quantified by Associated Regional and University Pathologists, Inc. (ARUP; Salt Lake City, UT).
A 48-year-old white man was found unresponsive in his home with an empty vial of U100 NPH insulin and a "probable" needle puncture in the left antecubital fossa. He was transported to a local hospital, where he remained comatose. At admission his blood glucose was 15 mg/dl (normal range: 70-110 mg/dl). Narcan did not provide an alteration in mental status, nor did six ampules of a concentrated sugar solution (D50). A blood alcohol concentration was 0.098% and a barbiturate screen was negative. He had a history of suicide attempts (shotgun wound of the abdomen). He had been known to carry a suicide note in his pocket. The source of the insulin was never determined. He remained in a persistent vegetative state, and during the subsequent 8 years he was transferred to numerous health care facilities before he finally died of pneumonia. An autopsy was not performed. The cause of death was recorded as complications of nonprescribed insulin injection, and the manner of death was suicide.
A relative found a 36-year-old white woman unresponsive in her bed in her home. On a bedside table was a bottle of insulin and a syringe. She was taken to a local hospital, and her blood glucose level was found to be 3 mg/dl (normal range: 70-110 mg/dl). Drugs of abuse, alcohol, salicylates, and acetaminophen were not detected in samples take at admission. Two days after admission, support was withdrawn because of brain death. An autopsy was not performed. The cause of death was recorded as insulin overdose, and the manner of death was suicide. This was based on the scene investigation, which included a statement made by the decedent several weeks earlier, "If I ever kill myself, it will be with insulin." The source of the insulin was not determined.
A 24-year-old white man was found in a school-yard, displaying seizure activity. He was transported to a local hospital, but was pronounced dead on arrival. No antemortem laboratory tests were performed. According to friends, he had been depressed during the past 2 weeks and had mentioned suicide. It was reported that one of his friends had witnessed him inject insulin several hours before he was found unresponsive. At autopsy, two recent needle punctures were on the left arm and no significant natural disease was identified. The pancreas was within normal limits. The blood alcohol level was 0.121%, and metabolites of marijuana were detected. Blood and vitreous were collected for insulin testing. Vitreous insulin was 31 μU/ml and the vitreous glucose was 26 mg/dl. Unfortunately, the blood specimen was determined to be inadequate because of hemolysis and insufficient quantity. The cause of death was recorded as consistent with hypoglycemia and hyperinsulinism, and the manner of death was suicide. The source of the insulin was not determined.
A female roommate found a 44-year-old white man dead on the floor near the couch in the late afternoon. The roommate reported that the decedent had injected himself with her insulin the night before. She checked his blood glucose with her fingerstick monitor and the level was 26 mg/dl. He drank some "sugar water" and went to bed. She had seen him alive the next morning before her leaving for work. When law enforcement officials arrived, she reported a vial of Novolin 70/30, a syringe, and 12 capsules of 25 mg diphenhydramine were missing from her medicine cabinet. She also reported that he had been depressed over the recent death of his wife. At autopsy, there were no visible injection sites. There were signs of chronic alcohol abuse including moderate hepatic steatosis with bridging fibrosis and pancreatic fibrosis. There were no gross or histologic signs of a pancreatic tumor. Drugs of abuse, diphenhydramine, and alcohol were not detected in postmortem specimens. Femoral blood insulin and C peptide were 840 μU/ml and 0.5 ng/ml, respectively. The cause of death was recorded as insulin intoxication and the manner of death was suicide.
The four cases presented in this report demonstrate a wide range of scenarios of suicide in nondiabetic individuals where the cause of death was insulin intoxication. Scene investigation and interviews allowed the emergency room personnel and the pathologists to consider exogenous insulin as the etiology of the hypoglycemia and as the eventual cause of death. In three of the cases, insulin was found at the scene. In the other case, the decedent had been observed injecting insulin before death.
In previous reports of suicidal insulin overdose in nondiabetic individuals, the source of the insulin was a mother (7), an ex-wife (8), a cohabitant (9), a husband (10), and, presumably, a hospital pharmacy (11). Kernbach-Wighton and Puschel report a series of 12 insulin-related deaths in which three decedents were not diabetic, and although not all of the deaths were discussed in detail, two decedents were relatives of diabetics and four were medical personnel (6). In this same report, details of one case of a nondiabetic male nurse were presented; however, the source of the insulin was not revealed (6). Although usually considered in relation to suicide, insulin has been reported as an agent of homicide (12-14) and in the case presented by Beastall et al. (9), the cohabitant was charged secondary to involvement (teaching the decedent how to inject insulin).
At autopsy, the external examination is usually unremarkable. In one of our cases (case 1), the emergency room staff reported a "probable" needle puncture site and in another (case 3), two needle punctures were noted at autopsy. There have been reports of insulin detection at injection sites via biochemical or histochemical methods (1,13,15,16). If the injection site can be found, these methods can be very helpful; however, most insulin needles are high gauge and the definitive injection site can be difficult, if not impossible, to find. In case 3, no ancillary tests of the injection sites were performed. Internal examination is also unremarkable because there are no anatomic findings that can be directly related to insulin toxicity. A thorough examination of the pancreas must be performed to rule out an insulin-secreting tumor (insulinoma). In the two cases in which an autopsy was performed, there was no gross or microscopic evidence of any type of pancreatic tumor.
Insulin is a peptide hormone secreted by the pancreatic beta cells in response to hyperglycemia (17). Active insulin consists of two peptide chains (A and B chains) that are linked by two disulfide bridges. Before secretion, the proinsulin molecule (inactive form) exists as one long chain with a connecting peptide (C peptide) between the A and B chains. Enzymes within the beta cell secretory granules cleave the C peptide before secretion; therefore, when the stimulus for secretion arrives, the granules release both active insulin and C peptide. In living persons, the plasma half-life of insulin is 5 to 8 minutes (17,18), and the plasma half-life of C peptide is 10 to 20 minutes (18). Insulin is metabolized by the liver, whereas C peptide is removed from the circulation by the kidneys (19,20). Furthermore, because both insulin and C peptide are secreted in equimolar concentrations, the ratio of insulin to C peptide should be near 1.0 (21) or even slightly higher because C peptide has the longer serum half-life. (For those who receive their laboratory results in conventional units [microunits/milliliter for insulin and nanograms/milliliter for C peptide], the ratio becomes 47.17 μU/ng.)
Once insulin has been considered as a potential cause of death, collection of the appropriate specimens is extremely important because insulin degrades rapidly at room temperature or higher (14,15). Sodium fluoride (15), serum separator tubes (22), and EDTA (22) are appropriate preservatives for blood specimens. Peripheral (femoral or iliac) blood is the best specimen for postmortem detection of insulin, because heart blood, especially the right-side heart blood, may show concentrations much higher than peripheral specimens (11,23). Bile has been used as a specimen in a patient admitted to the hospital for 9 days after insulin injection (8). In that case, insulin levels from admission were available for comparison because further studies of bile insulin levels in both diabetic and nondiabetic patients are quite variable (8). Tissue samples of liver, brain, or kidney are not recommended for insulin levels (1).
Once the specimen has been collected, storage becomes a major issue, because few medical examiners' offices have the equipment necessary to detect insulin. Haibach et al. stated that insulin is fairly stable at 4°C and that 80% to 95% of insulin is recoverable after 6 days at 4°C. However, insulin is nearly completely degraded during 4 to 5 days at 37°C (14). It is recommended that the serum be separated from the red blood cells as soon as possible and then refrigerated or, preferably, frozen. Animal versus human insulin can be determined via high-pressure liquid chromatography (3). Human recombinant insulin is the mainstay for current therapy, so species source determination may not be as important. A recent report has attempted to establish a method for quantification of insulin from dried blood spots. Dowlati et al. concluded that the dried blood spots were useful for field studies; however, they recommended that abnormal values be followed up with venous plasma samples (24), so this may be of little use for medical examiners.
Quantitation of C peptide, a cleavage product of endogenous proinsulin, is essential for the accurate interpretation of insulin levels. C peptide is not present in exogenous insulin (human or animal); thus, a low C-peptide level and a high insulin level suggest an exogenous source of insulin. Collection and storage of blood for C-peptide analysis is extremely important because C peptide is degraded in serum or improperly collected plasma (3). C peptide is more stable than insulin in postmortem blood (9). ARUP recommends that samples for C peptide be collected in serum separator tubes; heparinized samples are acceptable, and EDTA plasma samples are unacceptable (22). In contrast, Given et al. state that samples for C peptide should be collected in chilled tubes containing EDTA and trasylol (3). Both recommend that the plasma fraction should be separated and frozen within hours of collection (3,22). Coe reported that insulin and C peptide only minimally penetrate the blood-vitreous barrier (23). Of interest, in one of the cases presented, insulin was detected in vitreous. Unfortunately, a blood level was not available.
Patel provided an in-depth review of postmortem insulin detection and a table of values for normal fasting (5-75 μU/ml) versus fatal (800-3200 μU/ml) insulin concentrations (11). Some advocate measuring both free and total insulin levels to distinguish between free insulin in serum and the amount bound to insulin-specific autoantibodies (11,25). In the case of a suicide or a homicide, the insulin levels are likely to be so high, the difference between free and total is not likely to be important. In nondiabetic patients, free insulin levels most likely represent total amounts. In diabetic patients, it is recommended that both free and bound insulin should be measured because the insulin-specific antibodies can interfere with assay results and interpretation (25).
In summary, evaluation of the insulin-related death starts with the scene investigation. Even if there is only a slight suspicion of insulin overdose, the peripheral blood should be collected and the plasma fraction separated and frozen as soon as possible. As reported above, different laboratories may require different preservatives, so it is imperative to contact the laboratory that will be performing the analysis. With a greater awareness of proper collection and storage of blood, the number of cases in which insulin is implicated as a cause of death will most likely increase.
Acknowledgment: I thank Marcus Nashelsky, M.D., for his editorial assistance.
1. Dickson SJ, Cairns ER. The isolation and quantitation of insulin in post-mortem specimens-a case report. Forensic Sci
2. Critchley JAJH, Proudfoot AT, Boyd SG, Campbell IW, Brown NS, Gordon A. Deaths and paradoxes after intentional insulin overdosage [letter]. BMJ
3. Given BD, Ostrega DM, Polonsky KS, Baldwin D, Kelley D, Kelley RI, Rubenstein AH. Hypoglycemia due to surreptitious injection of insulin. Identification of insulin species by high-performance liquid chromatography. Diabetes Care
4. Martin FIR, Hansen N, Warne GL. Attempted suicide by insulin overdose in insulin-requiring diabetics. Med J
5. Cooper AJ. Attempted suicide using insulin by a non diabetic: a case study demonstrating the acute and chronic consequences of profound hypoglycemia. Can J Psychiatry
6. Kernbach-Wighton G, Puschel K. On the phenomenology of lethal applications of insulin. Forensic Sci Int
7. Price DE. Some observations on another case of insulin poisoning, with complications. Med Sci Law
8. Sturner WQ, Putnam RS. Suicidal insulin poisoning with nine day survival: recovery in bile at autopsy by radioimmunoassay. J Forensic Sci
9. Beastall GH, Gibson IH, Martin J. Successful suicide by insulin injection in a non-diabetic. Med Sci Law
10. Lutz R, Pedal I, Wetzel C, Mattern R. Insulin injection sites: morphology and immunohistochemistry. Forensic Sci Int
11. Patel F. Fatal self-induced hyperinsulinaemia: a delayed post-mortem analytical detection. Med Sci Law
12. Bauman WA, Yalow RS. Insulin as a lethal weapon. J Forensic Sci
13. Hood I, Mirchandani H, Monforte J, Stacer W. Immunohistochemical demonstration of homicidal insulin injection site. Arch Pathol Lab Med
14. Haibach H, Dix JD, Shah JH. Homicide by insulin injection. J Forensic Sci
15. Fletcher SM. Insulin. A forensic primer. J Forensic Sci Soc
16. Phillips AP, Webb B, Curry AS. The detection of insulin in postmortem tissues. J Forensic Sci
17. In Berne RM, Levy MN, Koeppen BM, Stanton BA. Physiology.
4th ed. St. Louis: CV Mosby, 1998.
18. Matthews DR, Rudenski AS, Burnett MA, Darling P, Turner RC. The half life of endogenous insulin and c-peptide in man assessed by somatostatin suppression. Clin Endocrinol
19. Johnson DG, Alberti KGMM, Faber OK, Binder C. Hyperinsulinemia of hepatic cirrhosis. Lancet
20. Kuzuya T, Matsuda A, Sakamoto Y, Tanabshi S, Kajinuma H. C-peptide immunoreactivity (CPR) in urine. Diabetes
21. Lebowitz MR, Blumenthal SA. The molar ratio of insulin to c-peptide. Arch Intern Med
22. Associated Regional and University Pathologists, Inc. web page@http://www.arup-lab.com.
23. Coe JI. Postmortem chemistry update. Emphasis on forensic application. Am J Forensic Med Pathol
24. Dowlati B, Dunhardt PA, Smith MM, Shaheb S, Stuart CA. Quantification of insulin in dried blood spots. J Lab Clin Med
25. Gennaro WD, Van Norman JD. Quantitation of free, total and antibody bound insulin in insulin treated diabetics. Clin Chem
© 2000 Lippincott Williams & Wilkins, Inc.