patterns of electrolytes in post mortem analysis of vitreous humour (see Coe 1993)
Dehydration pattern (e.g. salt consumption, water deprivation, and water loss in diabetes mellitus)
- Raised sodium (greater than 150 – 165mmol/L)
- Raised chloride (greater than 125 – 140mmol/L)
- Moderately raised urea (greater than 40 – 100mg/dL)
- Reduced sodium
- Reduced chloride
- Raised potassium (greater than 20mmol/L)
Low salt/hypotonic pattern (e.g. cystic fibrosis, fluid loss from the gastrointestinal tract, fatty liver, cirrhosis, and polydipsia).
- Reduced sodium (less than 130mmol/L)
- Reduced chloride
- Relatively reduced potassium (less than 15mmol/L)
- Raised urea
- Raised creatinine
- No substantial increase in sodium or chloride
(Expected vitreous electrolytes at post-mortem: sodium 140 – 145mmol/L (Coe 1993) or 150 – 155mmol/L (Dolinak 2005); chloride 115 – 125 mmol/L (Coe 1993); urea 10 – 15mg/dL (Coe 1993) or 25mg/dL (Dolinak 2005)).
Madea and Musshoff (2007) highlighted problems with using vitreous humour for post-mortem biochemical analysis; the ante-mortem values for the deceased individual were not known, different analytical methods were used on vitreous humour than for serum, and the equilibration rates were unknown.
The following notes regarding post-mortem biochemical analysis can be found in Coe (1993):
- Urea nitrogen – post-mortem serum levels are stable; average levels in those with no renal disease = 47.4mg/dL (16.9mmol/L), and in cases of sudden death = 13 – 15.5mg/dL (4.64 – 5.5mmol/L); vitreous humour levels apparently mirror serum levels; creatinine is stable and reflects serum levels
- Sodium – levels fall after death in serum, but are stable in vitreous humour in the early post-mortem period
- Chloride – serum levels fall after death, whilst vitreous humour levels fall slightly after death
- Potassium – very high in serum after death, but increases in the vitreous humour in a “linear fashion”, although the post-mortem level is affected by sampling and testing methods, body temperature, urea retention, and a more rapid increase after death in infants. The potassium level in vitreous humour is also erratic in those who died after chronic illness. Attempts at relating vitreous humour potassium concentration to post-mortem interval have been described (e.g. Madea in Henssge et al 2002), with the following equation being produced: post-mortem interval (in hours) = 5.26 x potassium concentration (in millimoles per litre) minus 30.9 (+ or – 20 hours in the first 100 hours post-mortem
notes on post mortem biochemistry (see Forrest 1993)
- Vitreous lactate – might be raised following prolonged agonal period
- Hypoxanthine – might be raised following prolonged agonal hypoxia
- Serum cholesterol and lipoprotein electrophoresis – analysis affected by haemolysis
- Cortisol – very high in stressful agonal periods, and very low in adrenal hypofunction
- Urine catecholamines – possibly high in hypothermia
- Vitreous amylase – occasionally raised in hypothermia
converting between units
- Millimoles per litre (mmol/L) of glucose to milligrams per decilitre (mg/dL) = times x 18
- Milligrams per decilitre (mg/dL) of glucose to millimoles per litre (mmol/L) = divide by 18 or times 0.055
- Millimoles per litre (mmol/L) to milligrams per decilitre (mg/dL) = times x 10.4
Diabetes mellitus classically presents with polydypsia, polyuria, polyphagia, and weight loss, although in some individuals (for example one third of children), diabetic coma is the first presentation.
Diabetes mellitus is a metabolic disorder characterised by hyperglycaemia and a failure to a greater or lesser extent to secrete insulin.
“Dead in bed syndrome” and type 1 diabetes mellitus
Start et al (2007) state that nocturnal hypoglycaemia is common in individuals with type 1 diabetes mellitus, and may account for individuals with diabetes who are unexpectedly found dead in bed.
Proposed mechanisms include the effects of an undetected autonomic neuropathy, with sympathetic overdrive leading to arrhythmias, but many of the individuals who are found unexpectedly dead in bed with diabetes are less than 40 years of age, with uncomplicated diabetes, so this mechanism is thought to be unlikely. An alternative mechanism is some disturbance of the electrophysiology due to hypoglycaemia, leading to non-stained ventricular tachycardia, and possibly an increase in the QT interval. It is also possible that hypokalaemia played a role, or that the individuals who die suddenly in this scenario have actually died from a channelopathy.
post mortem glucose analysis
- post mortem blood glucose levels fluctuate unpredictably, and there is a variable decrease in glucose after death, therefore hypoglycaemia cannot be diagnosed accurately from a post mortem analysis of vitreous humour glucose.
- post mortem blood glucose levels of greater than 500mg/dL (27.8mmol/L) have been found in 10% of non-diabetics, and blood glucose levels may be elevated post mortem in individuals with congestive cardiac failure, and following electrocution, asphyxia, terminal stress and cardiopulmonary resuscitation. Coe (1993) believes that the diagnosis of diabetes mellitus should not be made on post mortem glucose analysis alone
- Post mortem glycolysis has little effect on the raised levels of glucose in vitreous humour in uncontrolled diabetes mellitus and, even where non-diabetics had a raised blood glucose level at post mortem, none had a corresponding vitreous humour glucose level of greater than 100mg/dL (5.5mmol/L). Coe (1993) never found a vitreous humour glucose level of greater than 200mg/dL (11.1mmol/L) except in cases of diabetes mellitus
- Glucose can, however, be synthesised in vitreous humour in decomposed/embalmed bodies, and elevated vitreous humour glucose levels have been measured in cases of hypothermia (60 – 80mg/dL (3.3 – 10mmol/L))
An increased metabolism of free fatty acids results in ketone bodies and acidosis in diabetes mellitus, and the combination of hyperglycaemia and raised ketone levels is diagnostic of diabetic ketoacidosis. This should be compared with raised ketones without corresponding rising glucose in alcoholic ketoacidosis.
In diabetic ketoacidosis electrolyte depletion is exacerbated by osmotic diuresis. Individuals suffering from diabetic ketoacidosis may have an acetone smell to their breath, and there may be associated kussmaul breathing (deep sighing).
post mortem diagnosis of diabetic ketoacidosis
- greater than or equal to 300mg/dL (16.5mmol/L), with an average of 736mg/dL (40.5mmol/L)
- Pounder et al (1998) identified total blood ketones post mortem of 17.4 – 87.5mmol/L in diabetics with diabetic ketoacidosis, and DiMaio et al (1977) found blood acetone levels of 14.5 – 74.95mg/dL in diabetics with ketoacidosis (compared with a blood acetone level of less than 0.17mg/dL in individuals without diabetes)
Post mortem beta-hydroxybutyrate levels of 2.3 – 37.8mmol/L were present in diabetics with ketoacidosis by Iten and Meier (2000), compared with a “normal” level of less than 0.5mmol/L in individuals without diabetic ketoacidosis. Fasting levels of beta-hydroxybutyrate were identified in post mortem blood samples by Thomsen (1995) of 0.03 – 0.65mmol/L following fasting.
Iten and Meier (2000) are of the opinion that post mortem blood beta-hydroxybutyrate levels of between 0.5 – 2.5mmol/L represents an “elevated” level, whilst greater than 2.5mmol/L represents a “pathological” level (diabetic or alcoholic ketoacidosis).
Local biochemistry laboratories tend to apply their own reference ranges for chemical values found in blood etc. The University Hospital of Wales (UHW), Cardiff, UK, for example, uses the following reference range for blood beta-hydroxybutyrate: 0 – 0.44mmol/L.
non-ketotic diabetic coma
DiMaio et al (1977) identified the characteristics of individuals with non-ketotic diabetic coma as having a very high blood glucose level (average 1949mg/dL (107mmol/L)), in a setting of gross dehydration, raised sodium and potassium levels, and a blood acetone level of 5.81mg/dL.
post mortem vitreous humour analysis and diabetes
DiMaio et al (1977) recognised a vitreous humour glucose level of greater than 200mg/dL (greater than 11mmol/L) as being “diagnostic”, whilst Iten and Meier (2000) recognised a concentration of 11.6 – 63.2mmol/L as being found in vitreous humour from diabetics (compared with a vitreous humour concentration of 3.9 – 5.8mmol/L in “normal” individuals).
Osuna et al considered a vitreous humour glucose concentration of 42.5mg/dL as being 87% specific for diabetes mellitus.
Denmark (1993), and Iten and Meier (2000) recognised a beta-hydroxybutyrate concentration in vitreous humour post mortem of between 1.8 and 8.2mmol/L (18.72 – 85.28mg/dL) as being representative of ketoacidosis, compared with control concentrations of less than 0.96mmol/L (less than 9.98mg/dL).
post mortem analysis of insulin
The absence of insulin in post mortem blood samples was not considered to be of importance.
The normal plasma insulin level is said to be 6 – 26mU/ml (whilst some authors say that the post-mortem insulin level should be at a maximum of 0.07mU/ml).
suspected exogenous administration
Birkinshaw et al (1958) described the case of R – v – Barlow, in which the first murder charge was brought for exogenous insulin administration.
When considering whether the death is related to exogenous insulin administration, one must exclude an islet cell tumour (pancreatic endocrine neoplasia for example insulinoma (B-cell tumour)).
Endogenous insulin production (for example from tumours) is accompanied by a corresponding equivalent increase in c-peptide (which is split from proinsulin).
Exogenous insulin administration, which leads to glucose being taken up in tissues, leading to coma, is not associated with a rise in the c-peptide levels. Surreptitious administration of insulin has been shown by Haibach et al (1987) to lead to blood insulin levels of 12 – 40,600mU/ml.
Secretogogue drugs, for example tolbutamide and chlorpropramide, cause there to be a rise in endogenous insulin production, with a corresponding rise in c-peptide.
One can’t calculate the dose of insulin administered exogenously from post mortem blood analysis.
Post mortem analysis could include measuring antibodies to animal insulins, but one should then be aware of previous exposure effects in that individual.
The investigation of a suspect exogenous insulin-administration death would need to consider whether the deceased had access to insulin , for example health professions, or those related to somebody with diabetes.
No post mortem correlation has been found between insulin dosages and post mortem values, or the time intervals of administration (Kernbach-Wighton and Puschel 1998).
No defined “fatal” post mortem levels of insulin in blood have been identified, but concentrations measured in fatal cases of suspected exogenous insulin administration have been recorded as 12 – 40,00mU/ml (with a corresponding blood glucose of 0.25mmol/L, and a c-peptide level of less than 0.1 – 4.4ng/ml). Others have identified levels of between 1 – 155µmol/L (mean 49µmol/L).
Insulin to C-peptide ratio
Exogenous insulin administration leads to a rise in the insulin to c-peptide ratio (which is normally 0.1 – 0.47, and “never” greater than 1). Of note, the insulin to c-peptide ratio is normal in those with insulinomas.
- When the c-peptide level is low (i.e. 55 – 111pmol/L) the insulin to c-peptide ratio is high (0.545 +/- 0.21)
- When the c-peptide level is high (138 – 221pmol/L), the insulin to c-peptide ratio is low (0.436 +/- 0.26)
Of note, the c-peptide levels fall after approximately 24 hours post mortem, and the insulin to c-peptide ratio increases.
In the living, the clinical management of individuals with diabetes mellitus includes the aim to reduce the HbA1c to below 7%.
The HbA1c protein analysis aims to identify the long-term glucose control in individuals with diabetes.
Post mortem blood analysis of non-diabetics has shown HbA1c to be between 4 and 6% (Valenzuela 1998), and between 3.25 – 6.25% (Goulle et al 2002).
Valenzuela (1988) found HbA1c at post mortem to be 6.5 – 10.7% in diabetics.
Kernbach- Wighton and Puschel (1998) found post mortem HbA1c in exogenous insulin administration cases to be between 8.1 – 10.7%, and an HbA1c of greater than 12% was identified in diabetic comas.
Nathan’s formula for estimating mean blood glucose concentrations, using HbA1c, is stated as HbA1c x 1.85 – 4.78.
Insulin can be detected in blood by RIA (and antibodies to determine the type of insulin can interfere with this. Post mortem haemolysis can also interfere with insulin level estimation.
This phenomenon usually occurs after abstinence from alcohol, and in a setting of reduced food intake, where free fatty acids have built up, and cannot be oxidised, leading to increased ketone production.
There is no elevation of glucose in alcoholic ketoacidosis as there is in diabetic ketoacidosis.
Excess alcohol caused a rise in lactate and ketogenesis and, in the setting of concomitant malnourishment, there is a rise in liver glycogen utilisation, leading to increased lipolysis and ketogenesis. Vomiting also exacerbates the development of ketoacidosis.
Post mortem biochemical analysis and alcoholic ketoacidosis:
- Total ketones greater than 10mmol/L (Pounder et al (1998)); greater than 0.5mmol/L (Thompson (1995)).
- Beta-hydroxybutyrate 0.7 – 20.5mmol/L (average 5.5 +/- 3.6mmol/L) (Thompson (1995)); 1.26 – 47mmol/L (greater than 2.5mmol/L = abnormal) (Iten & Meier (2000)).
vitreous humour ketones
- Total vitreous humour ketones greater than 5mmol/L (Pounder et al 1998)
- Vitreous humour beta-hydroxybutyrate 1.83 – 8.17mmol/L (Denmark 1993)
Mechanisms of sudden death in alcoholics with fatty liver but no post mortem blood alcohol detected
- Subtle infection
- Hepatorenal metabolic alterations
- Coe JI. Postmortem chemistry update – emphasis on forensic application. American Journal of Forensic Medicine and Pathology 1993; 14:91-117
- Dolinak D. Toxicology. Chapter 21 In: Forensic pathology: principles and practice. Dolinak D, Matshes EW, Lew EO (Eds), Elsevier/ Academic Press, London 2005
- Madea B, Musshoff F. Postmortem biochemistry. Forensic Science International 2007; 165:165-171
- Madea B, Henssge C. Chapter 4 In: The estimation of the time since death in the early post-mortem period, 2nd Edition, Henssge C, Knight B, Krompecher T et al (Eds), Arnold Publishing UK 2002
- Pounder DJ, Stevenson RJ, Taylor KK. Alcoholic ketoacidosis at autopsy. Journal of Forensic Sciences 1998; 43:812-816
- DiMaio VJM, Sturner WQ, Coe JI. Sudden and unexpected deaths after the acute onset of diabetes mellitus. Journal of Forensic Sciences 1977; 22:147-151 [quoting Sulway JJ, Trotter W, Trotter MD, Malins JM. Acetone in uncontrolled diabetes. Postgraduate Medical Journal 1971 June Supplement pp.382-387]
- Iten PX, Meier M. Beta-hydroxybutyric acid – an indicator for an alcoholic ketoacidosis as cause of death in deceased alcohol abusers. Journal of Forensic Sciences 2000; 45:624-632
- Thomsen JL, Felby S, Theilade P, Nielsen E. Alcoholic ketoacidosis as a cause of death in forensic cases. Forensic Science International 1995; 75:163-171 [blood BHB in fasting subjects quoted from Williamson DH, Mellanby, Krebs HA. Enzymic determination of D(-)-β-hydroxybutyric acid and acetoacetic acid in blood. Biochemistry Journal 1962; 82:90-96]
- Osuna E, Vivero G, Conejero J et al. Postmortem vitreous humour β-hydroxybutyrate: its utility for the post-mortem interpretation of diabetes mellitus. Forensic Science International 2005; 153:189-195
- Denmark LN. The investigation of beta-hydroxybutyrate as a marker for sudden death due to hypoglycaemia in alcoholics. Forensic Science International 1993; 62:225-232
- Birkinshaw VJ, Gurd MR, Randall SS et al. Investigations in a case of murder by insulin poisoning. British Medical Journal 1958; August 23 pp.463-468
- Haibach H, Dix JD, Shah JH. Homicide by insulin administration. Journal of Forensic Sciences 1987; 32:208-216
- Kernbach-Wighton G, Puschel K. On the phenomenology of lethal applications of insulin. Forensic Science International 1998; 93:61-73
- Valenzuela A. Postmortem diagnosis of diabetes mellitus. Quantitation of fructosamine and glycated haemoglobin. Forensic Science International 1988; 38:203-208
- Goulle J-P, Lacroix C, Bouige D. Glycated haemoglobin: a useful post-mortem reference marker in determining diabetes. Forensic Science International 2002; 128:44-49
- Start RD, Barber C, Kaschula ROC, Robinson RTCE. The ‘dead in bed syndrome’ – a cause of sudden death in Type 1 diabetes mellitus. Histopathology 2007; 51:843-845