Forensic Science International 215 (2012) 189198

Contents lists available at ScienceDirect

Forensic Science International

journal homepage: www.elsevier.com/locate/forsciint

Microbial ethanol production: Experimental study and multivariate evaluation

Vassiliki A. Boumba a,*, Vangelis Economou b, Nikolaos Kourkoumelis c, Panagiota Gousia b,
Chrissanthy Papadopoulou b, Theodore Vougiouklakis a
a

Department of Forensic Medicine & Toxicology, Medical School, University of Ioannina, 45110 Ioaninna, Greece
Department of Microbiology, Medical School, University of Ioannina, 45110 Ioaninna, Greece
c
Department of Medical Physics, Medical School, University of Ioannina, 45110 Ioaninna, Greece
b

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 30 September 2010
Received in revised form 3 March 2011
Accepted 7 March 2011
Available online 5 April 2011

Ethanol can be produced from all the postmortem available substrates, though with higher rates and
yields from carbohydrates, during the early stages of putrefaction. The so-called higher alcohols (1propanol, isobutanol, 2-methyl-1-butanol and 3-methyl-2-butanol) and 1-butanol could be produced,
from all the available postmortem substrates. However, a quantitative relationship between the
produced ethanol and the potentially produced other alcohols is still missing.
The objective of this study was the development of a simple, mathematical model which could be able
to approximate the microbial produced ethanol in correlation with other produced alcohols. The selected
bacterial species included two Gram+ spore-forming anaerobic bacteria and two (one Gram+ one Gram-)
aerobic/facultative anaerobic bacteria, all being common commensals of the digestive tract and common
colonizers of the corpse. The selected bacterial strains, Escherichia coli, Clostridium perfrigens, Clostridium
sporogenes and Enterococcus faecalis, were cultured separately at 25 8C, for 30 days, under controlled
anaerobic conditions. The produced ethanol and the previously referred alcohols were determined in the
culture medium in 24 h intervals. Using partial least squares (PLS) regression, the estimation of the
relevance score for the available descriptors established the statistical model to assess the ethanol
concentration produced by each studied microbe. E. coli, C. perfrigens, and C. sporogenes produced
different patterns of ethanol and other alcohols, while E. faecalis produced negligible amounts of ethanol
and higher alcohols. In constructing the mathematical models to predict the produced ethanol, 1propanol, 1-butanol, and isobutanol were signicant for C. perfrigens and C. sporogenes, while 1-butanol,
1-propanol, and methyl-butanol were signicant for E. coli. The applicability of these models was tested
in microbial, anaerobic cultures of normal human blood and plasma at 25 8C. The results indicate that
factors such as the type of microbe species, the glucose content and the medium composition apparently
affect the procedure of microbial ethanol, and other alcohols production. However, the models can be
applied with acceptable accuracy and they show potential for application in real postmortem cases.
2011 Elsevier Ireland Ltd. All rights reserved.

Keywords:
Post-mortem blood
Ethanol
Microbial ethanol production
Multivariate statistics
Fermentation
Volatiles
Higher alcohols
Biomarker(s)

1. Introduction
The microbial formation of ethanol is a well recognized
complication in the proper interpretation of postmortem ethanol
analysis results [13]. Many microorganisms potentially present in
a dead body are capable of ethanol production. At least 58 species
of bacteria, 17 species of yeasts and 24 species of molds can
produce ethanol as well as other volatiles through various
biosynthetic pathways [1,4,5].

48th Annual Meeting of the International Association of Forensic Toxicologists

(TIAFT). Joint meeting with the society of Toxicological and Forensic Chemistry
(GTFCh).
* Corresponding author. Tel.: +30 26510 07724; fax: +30 26510 07857.
E-mail addresses: vboumba@cc.uoi.gr, metsbou@yahoo.g (V.A. Boumba).

0379-0738/$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.forsciint.2011.03.003

However, only a limited number of microbial species have been

reported to produce ethanol in experimental studies [611]. From
the available relevant literature the microbial species studied so far
are: Candida tropicalis in blood [6], Candida albicans in blood [6,7]
and in urine [8,9], Candida tropicana in blood [6] and in urine [8],
Candida parapsilosis in blood [6] and in urine [9], Corynebacterium
sp. in blood [6], Lactococcus garviae in blood [10], Escherichia coli in
blood [6] and in urine [8,9], Candida glabrata in urine [11],
Enterococcus and Klebsiella oxytoca in urine [8], and Klebsiella
pneumoniae and Proteus mirabilis in urine [9].
In particular, the species C. albicans, C. tropicalis, C. tropicana, C.
glabrata, C. paralsilosis, E. coli, K. oxytoca, Corynebacterium sp.,
Enterococcus and L. garviae have been identied in postmortem
blood and urine [6,8,10,11], while only C. albicans, C. parapsilosis, K.
pneumoniae, E. coli, Proteus mirabilis, and S. cerevisiae have been
experimentally inoculated in human blood or urine [7,9,12].

190

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

Candida spp. have been the most studied species, followed by E.

coli.
It has been suggested that ethanol can be produced from all the
available substrates in the postmortem, although with higher rates
and yields from carbohydrates, during the early stages of
putrefaction [1,4,5]. The formation by bacteria of the so-called
higher alcohols (1-propanol, isobutanol, 3-methyl-2-butanol and
2-methyl-1-butanol) from their relevant alpha-ketoacids (using
aminoacids as substrates) is supported by several data (reviewed
in Ref. [5]). Clostridia could produce 1-butanol through the
butyrate and butanol-acetone fermentation pathway which
proceeds with the parallel production of ethanol. The higher
alcohols and 1-butanol could be produced simultaneously with
ethanol, from all the available postmortem substrates by branched
biochemical pathways [5]. Remarkable aspects of their formation
include the dependence on glucose metabolism, their appearance
coinciding with the formation of ethanol and the predominance of
1-propanol in most cases, although the quantitative pattern of
their formation is variable.
Among the volatiles that have been correlated to postmortem
ethanol formation, 1-propanol and 1-butanol appear to have a
supreme role. Specically, the concentration of 1-propanol has
been suggested to have a quantitative relationship with the
postmortem produced ethanol [1315] although this approach has
been questioned by other researchers [7]. The presence of 1butanol, on the other hand, has been considered as a qualitative
indicator of postmortem ethanol production in drowned persons
[16]. Yet a quantitative relationship between the produced ethanol
and the potentially produced other alcohols is still missing.
The objective of this study was the development of the simplest,
mathematical model which could estimate the microbial produced
ethanol in correlation with other produced alcohols. Bacterial
species, reported to be common corpses colonizers and ethanol
producers [4,5], were cultured under controlled anaerobic conditions and the produced ethanol and other alcohols were
determined by head space-gas chromatographyame ionization
detector (HS-GCFID). The results were analyzed by multiple
regression analysis and the correlation among the produced
alcohols concentrations was described by mathematical equations
(models). The applicability of these models was tested in microbial
cultures of normal human blood and plasma under laboratory
environment. Finally, the models were applied in post-mortem
cases and the main outcomes and limitations so far are discussed.
2. Materials and methods
2.1. Microbial anaerobic cultures
2.1.1. Bacterial strains
There were selected 4 bacterial species, including two Gram+ spore-forming
anaerobic bacteria and two (one Gram+ one Gram) aerobic/facultative anaerobic
bacteria, all being common commensals of the digestive tract, common colonizers
of the corpse and typical anaerobic and aerobic/facultative anaerobic species. The
bacterial strains used in this study were Clostridium perfrigens NCTC 6785,
Clostridium sporogenes NCTC 8594, Enterococcus faecalis ATCC 376 and Escherichia
coli ATCC 11303. All bacterial isolates were stored in 80 8C using the beads of the
Microbank system (ProLab Diagnostics, Canada). The strains were revived by
transferring 35 beads into 10 mL of Brain Heart Infusion (BHI) (CM0225, Oxoid)
and incubating at 37 8C for 1824 h. BHI is a very good alternative to blood, safer,
cheaper, enabling easier counting with a spectrophotometer and it is preferred by
our Microbiology Department in experimental studies. The obligate anaerobes C.
perfrigens and C. sporogenes were incubated under strict anaerobic conditions by
using a sterile parane oil overlay and the GENbag anaer incubation system (45 534,
BioMerieux). For E. faecalis and E. coli a sterile parane oil overlay was used.
Subsequently 100 mL from the incubated broth was inoculated in 10 mL of BHI and
incubated at 25 8C for 1824 h. Before inoculation the bacterial count was adjusted
at 108 cells/mL (optical density 0.605) with the use of a UV spectrophotometer
(Model 6305, Jenway, England) at a wavelength of 620 nm. For the appropriate
dilutions sterile BHI was added. Sterile glass tubes containing 10 mL of sterile BHI
were inoculated with the appropriate volume so as an initial concentration of
104 cells/mL to be obtained. A sterile parane oil overlay of 1.0 mL and the GENbag

anaer incubation system (45 534, BioMerieux) was used to obtain anaerobic
conditions. The inoculated tubes were incubated under anaerobic conditions at
25 8C and 4 8C for 30 days.
2.1.2. Bacterial enumeration
The bacterial counts were performed by the pour plate technique. Decimal
dilutions were performed with the use of tubes containing 9 mL of Maximum
Recovery Dilluent (02-510, Scharlau Chemie, Spain). From each dilution an
inoculum of 1 mL was added into two sterile petri dishes and 15 mL of melted
Plate Count Agar (PCA) (CM0325, Oxoid) at a temperature of 45 8C was poured. After
solidication, 5 mL of plate count agar were overlaid. The plates were inverted and
incubated for 48 h at 30 8C under anaerobic (C. perfrigens and C. sporogenes) or
aerobic (E. coli and E. faecalis) conditions. Plates with 30300 colonies were selected
and colonies were counted under a colony counter (Stuart SC6, Barloworld Scientic
Ltd., UK). The average from the two plates from the appropriate dilution was
calculated. The result was derived from the multiplication of the average colony
count with the dilution factor and was expressed as cfu/mL.
2.1.3. Microbial cultures in human blood and plasma products
Human whole blood from healthy individuals was collected in sterilized blood
tubes containing EDTA (Vacuette EDTAK3, 6 mL), and was centrifuged at 4000 rpm
for 5 min. The supernatant (plasma) was collected and 1 mL portions were placed
into 16 sterilized blood tubes (Vacuette EDTAK3, 3 mL). Each tube was inoculated
with the appropriate volume of bacterial broth of either E. coli, or C. perfrigens or C.
sporogenes in order to obtain an initial concentration of 104 cells/mL (equal to the
culture concentrations used to build the relevant microbial model). Samples were
incubated under anaerobic conditions by using a sterile parane oil overlay and the
GENbag anaer incubation system. At days 0, 1, 2, 3, 5, 7, 11 and 15 two tubes were
removed from the incubation system and alcohols concentrations were determined
by HS-GCFID. Also, two other series of 16 tubes each, containing 1 mL of plasma,
were spiked with the appropriate volume of a concentrated glucose solution (sterile
dextrose 5% (w/v), pyrogen free solution for intravenous infusion, DEMO Ltd.,
Greece) so as to obtain 200 mg/dL nal concentration. The initial glucose
concentration of the spiked plasma samples was chosen to be 200 mg/dL in order
to be the same with the glucose content of the BHI culture medium. Next the
samples were inoculated with E. coli or C. sporogenes and were tested for their
alcohol production capacity as previously described. The aforementioned procedure was also applied to two series of 16 tubes each, containing whole human blood
with EDTA, and the other whole human blood with citric ions, inoculated with C.
perfrigens. All the aforementioned tests in human whole blood and plasma products
were performed in parallel.
2.2. Volatiles determination
2.2.1. Chemicals and solutions
All chemicals were purchased in the highest possible purity and used without
any further purication. Ammonium sulfate, ethanol (99.7%), 1-propanol, 1butanol, isobutanol (methyl-1-propanol), 2-methyl-1-butanol, 3-methyl-2-butanol, and acetonitrile were purchased from Merck in the higher available purity
(Darmstadt, Germany). All aqueous solutions were prepared using double distilled
(DD) water, which was obtained using an Aquatron A400D (Bibby Sterilin,
Staffordshine, UK).
Stock aqueous standard solutions were prepared in concentration 4.00% (w/v) for
ethanol; 0.25% (w/v) for 1-butanol, isobutanol, 2-methyl-1-butanol, and 3-methyl2-butanol; and 0.40% (w/v) for 1-propanol. Acetonitrile was used as internal
standard for the determination of volatiles in aqueous solution of 100 mg/dL. The
above solutions were stored at 4 8C for up to six months. Working solutions were
prepared on a daily basis by mixing the appropriate volumes of the corresponding
stock solutions of each analyte and DD-water. Blood quality control (QC) samples
were prepared from stock solutions in three different concentrations for each
analyte.
Calibration solutions separately for each analyte, were prepared daily, by spiking
aliquots of blank human blood (whole human blood that was certied to be volatile
organic compound free) with the appropriate volume of working solution.
2.2.2. HS-GCFID procedure
GC analyses were performed on a Shimadzu GC 17A gas chromatograph
equipped with a SUPELCOWAXTM10 fused silica capillary column
(30 m  0.25 mm, lm thickness 0.25 mm) and with a FID. The GC was tted with
a Shimadzu AOC-5000 headspace-GC automated sample pretreatment and
injection system. The analyses were performed by slight modications to
previously published methods [1719]. The temperature of the injection port,
the column and the FID was 115 8C, 60 8C and 260 8C, respectively. The carrier gas
was helium with a ow rate of 0.7 mL/min with a constant pressure of 65 kPa. The
injection inlet was set in a split mode with a split ratio of 10:1
The samples were incubated at 50 8C for 8.0 min [18] prior to injection with an
agitation speed of 500 rpm. A 2.5 mL gas tight syringe was used heated at 105 8C. A
head space aliquot of 500 mL was sampled for analysis with a feel speed of 250 mL/s
and a pull up delay 500 ms. The injection speed was 1000 mL/s at a needle

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

191

Table 1
Validation of the HS-GCFID volatiles determination method.
Volatile

LOD (mg/dL)

LOQ (mg/dL)

Working range (mg/dL)

R2

Ethanol
1-Propanol
Isobutanol
1-Butanol
3-Methyl-2-butanol
3-Methyl-1-butanol

0.01
0.02
0.005
0.005
0.01
0.01

0.03
0.04
0.015
0.01
0.025
0.025

0.03400
0.0432.0
0.020.08
0.010.08
0.040.32
0.040.32

0.999
0.999
0.989
0.995
0.997
0.999

penetration depth of 20 mm. After injection syringe was ushed with helium for
1.5 min. Acquisition time was 20 min. A standard GCFID chromatogram is
provided as complimentary data (Fig. 1C).
In 10 mL HS vials containing 0.50 g ammonium sulfate were added 500 mL of the
calibration solution or of the culture medium or of blood culture and 500 mL of the
internal standard solution. The vials were then sealed with metal crimp caps tted
with silicon septa and were put to the HS auto sampler for analysis. Ammonium
sulfate was added to increase the ionic strength of the solution. Calibration curves
were constructed for each analyte in six concentration levels within the relevant
working concentration range (Table 1).

3.2. Method validation

1,E+08

In this study a HS-GCFID method has been optimized, in order

to determine simultaneously the concentrations of ethanol, 1propanol, 1-butanol, isobutanol, 2-methyl-1-butanol and 3methyl-2-butanol, in the culture medium and in human blood
samples. These volatiles were baseline separated from other
common determinants in volatile blood analysis (methanol,
acetaldehyde, 2-propanol, acetone, ethyl acetate, butanone, 2butanol). Baseline separation of 2-methyl-1-butanol (active-amyl
alcohol) and 3-methyl-2-butanol (isoamyl alcohol) is not feasible
on polar columns typically used for separation of alcohols [22].
Therefore, herein they are determined as mixture and referred as
methyl-butanol. The method was evaluated for selectivity, limit of
detection (LOD), limit of quantitation (LOQ), linearity, precision
and accuracy. Selectivity was accomplished by analyzing six
different blank samples and the matrix effect was evaluated. The
LODs and LOQs for each analyte were determined as the lowest
concentration of each analyte yielding a signal to noise ratio of at
least 3:1 and 10:1 respectively. LODs and LOQs are listed in Table 1.
Linearity was expressed by the correlation coefcient (R2) of the
regression line calculated by the method of least-squares with a
weighting factor of 1/x2. Correlation coefcient exceeded 0.999 for
each analyte (Table 1). Each calibrator was back calculated against
the total curve. Precision and accuracy of the method were
evaluated by analyzing three QC samples, one at the lower
concentration, one at the higher and one within the working range
of concentration of each analyte. Precision was expressed as the
relative standard deviation (% RSD) and was lower than 6.0% for all
analytes. Accuracy of the method was calculated as the percent
difference from the expected concentration (% Er) and ranged from
1.0% to 2.5%.
The applied methodology was based on previously reported
methods [1719] and met acceptable analytical criteria for the
quantitative determination of alcohols. Acetonitrile was used as
internal standard since it has similar physicochemical properties
with the determined volatiles; it is neither an ingredient of
alcoholic beverages nor a putrefactive product [5]; and it was also
used in ethanol analysis in the past [17]. The LODs and LOQs of our
method lie between the values of previous reported methods being
either lower [19] or higher [18,23].

1,E+06

3.3. Alcohols determination

1,E+04

Under the applied experimental conditions the clostridia

species and E. coli produced ethanol, 1-butanol and higher alcohols
in variable quantities (Tables 24). The higher amounts of ethanol
were produced from C. sporogenes, followed by E. coli and C.
perfrigens. The most abundant produced other alcohol was 1butanol for C. sporogenes and C. perfringens, followed from 1propanol and isobutanol, while for E. coli was 1-propanol. The less
abundant produced volatile were methyl-1-butanol for all the

2.3. Multivariate statistical analysis

Regression analysis was employed to model the correlation between the
microbial produced ethanol and the other higher alcohols. The experimental data
were analyzed with multiple linear regression (MLR) methods to model the
relationship between ethanol concentrations as the dependent variable being in
linear correlation with the concentrations of the other alcohols (independent
variables). Using the model built, we calculated each descriptors relevance to the
current model using partial least squares (PLS) regression based on the PLS1
algorithm. Since MLR is very sensitive to outliers, the dimensionality of some
models has been decreased by leaving out certain independent variables. This has
also permitted us to nd the simplest acceptable solution in each case.

3. Results and discussion

3.1. Microbial anaerobic cultures in BHI culture medium
C. perfrigens, C. sporogenes, E. coli and E. faecalis (former known
as Streptococcus faecalis) have been reported as being the main
colonizers in corpses and in parallel the main ethanol producers
[4]. C. perfrigens and C. sporogenes are Gram-positive, rod-shaped,
obligate anaerobic, spore-forming bacteria of the genus Clostridia
widely distributed in nature and also in the intestinal tract of
humans and other vertebrates, insects, and soil. E. coli is a Gram
negative, facultative anaerobic and non-sporulating, rod-shaped
bacterium that is commonly found in the lower intestine of warmblooded organisms (endotherms). E. faecalis is a Gram-positive
commensal bacterium, facultative anaerobic, inhabiting the
gastrointestinal tracts of humans and other mammals [20].
In Fig. 1 the growth curves for each bacterial strain during the
30 days of incubation at 25 8C are presented. By the fth day of

Bacterial counts (log CFU/ml)

incubation all strains reached the stationary phase of growth. At

4 8C no growth was observed for all studied bacterial strains, as it
was expected (not shown) [9,21].

1,E+14
1,E+12
1,E+10

1,E+02

10

15

20

25

30

Days
Clostridium sporogenes
Enterococcus faecalis

Clostridium perfringens
Escherichia coli

Fig. 1. Growth curves for each microbe under study showing the number of bacteria
during the incubation period of 30 days at 25 8C under anaerobic conditions.

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

192

Table 2
Volatiles concentration during the fermentation period of C. perfrigens in BHI culture medium at 25 8C.
Days

Ethanol (g/L)

1-Propanol (mg/dL)

1-Butanol (mg/dL)

Isobutanol (mg/dL)

Methyl-butanol (mg/dL)

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

0.01
0.01
0.02
0.06
0.05
0.08
0.07
0.11
0.13
0.09
0.12
0.13
0.14
0.12
0.16
0.14
0.13
0.14
0.13
0.14
0.14
0.14
0.15
0.14
0.15
0.14
0.15
0.15
0.15
0.16
0.15

0
0.02
0.04
0.08
0.15
0.2
0.26
0.31
0.32
0.39
0.40
0.46
0.50
0.51
0.52
0.48
0.42
0.55
0.51
0.52
0.53
0.51
0.55
0.63
0.65
0.55
0.64
0.60
0.68
0.69
0.70

0
0.04
0.08
0.16
0.44
0.57
0.78
0.72
0.99
0.72
0.86
1.08
1.35
1.02
1.51
1.18
1.21
1.26
1.23
1.27
1.31
1.25
1.30
1.20
1.16
0.98
1.52
1.47
1.32
1.90
1.33

0
0
0.01
0.01
0.03
0.04
0.05
0.07
0.08
0.07
0.07
0.10
0.11
0.10
0.11
0.11
0.11
0.11
0.10
0.09
0.10
0.11
0.11
0.10
0.13
0.13
0.14
0.13
0.12
0.12
0.15

0
0
0.02
0
0.03
0.03
0.05
0.04
0.06
0.04
0.06
0.07
0.07
0.09
0.09
0.07
0.05
0.04
0.07
0.07
0.06
0.06
0.07
0.08
0.07
0.07
0.08
0.09
0.08
0.10
0.08

studied microbes. E. faecalis produced negligible amounts of

ethanol and other alcohols indicating that under the applied
conditions the branched fermentation pathways producing
alcohols were unfavorable for this microbe. Therefore, the
statistical evaluation of the relevant results was practical only

for the species C. pergens, C. sporogenes and E. coli and impractical

for E. faecalis.
The higher alcohols are metabolic products of the fermentation
mainly of amino acids which are used by microbes as a nitrogen
source. The higher alcohols, under the applied experimental

Table 3
Volatiles concentration during the fermentation period of C. sporogenes in BHI culture medium at 25 8C.
Days

Ethanol (g/L)

1-Propanol (mg/dL)

1-Butanol (mg/dL)

Isobutanol (mg/dL)

Methyl-butanol (mg/dL)

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

0.00
0.10
0.73
0.76
0.85
0.87
0.85
0.84
0.78
0.71
0.79
0.81
0.84
0.75
0.77
0.81
0.78
0.81
0.79
0.76
0.77
0.78
0.76
0.66
0.70
0.70
0.73
0.76
0.73
0.67
0.72

0.04
2.42
2.67
4.59
5.07
5.73
5.67
5.82
6.20
5.06
6.55
6.59
6.68
6.67
7.10
8.16
7.50
8.12
7.81
7.62
7.67
8.55
8.21
8.20
8.47
8.64
8.69
8.88
9.36
10.5
10.2

0
0.43
0.70
2.07
2.31
3.56
3.25
3.72
3.81
3.22
4.77
4.81
5.19
6.35
6.45
7.59
6.98
8.85
8.25
8.74
8.00
9.20
8.87
9.35
9.70
10.2
10.5
9.57
11.1
10.2
11.9

0.00
0.70
0.72
1.53
1.76
2.58
2.30
2.56
2.40
2.24
3.03
3.12
3.18
3.36
3.51
4.36
4.57
5.07
4.33
3.99
4.10
4.43
4.20
4.35
4.67
5.38
4.82
4.75
4.90
5.50
6.10

0
0.07
0.10
0.18
0.22
0.24
0.23
0.25
0.21
0.20
0.34
0.32
0.32
0.39
0.38
0.44
0.36
0.57
0.52
0.53
0.52
0.51
0.55
0.52
0.55
0.55
0.55
0.56
0.55
0.55
0.57

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

193

Table 4
Volatiles concentration during the fermentation period of E. coli in BHI culture medium at 25 8C.
Days

Ethanol (g/L)

1-Propanol (mg/dL)

Isobutanol (mg/dL)

1-Butanol (mg/dL)

Methyl-butanol (mg/dL)

0
0.5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

0
0.18
0.36
0.42
0.43
0.44
0.46
0.45
0.44
0.46
0.48
0.47
0.47
0.46
0.50
0.45
0.49
0.49
0.52
0.46
0.46
0.50
0.51
0.56
0.46
0.53
0.49
0.53
0.50
0.52
0.49
0.50

0
0.06
0.09
0.11
0.18
0.24
0.34
0.37
0.38
0.43
0.46
0.45
0.51
0.54
0.62
0.68
0.66
0.72
0.66
0.69
0.76
0.70
0.79
0.81
1.02
0.92
0.92
0.95
0.91
0.86
0.86
0.88

0
0.01
0.02
0.02
0.02
0.02
0.03
0.03
0.03
0.03
0.02
0.03
0.02
0.03
0.03
0.02
0.02
0.02
0.03
0.02
0.03
0.04
0.03
0.02
0.08
0.07
0.07
0.05
0.02
0.07
0.09
0.10

0
0.01
0.04
0.07
0.08
0.09
0.09
0.09
0.10
0.11
0.09
0.10
0.10
0.12
0.09
0.11
0.10
0.12
0.10
0.10
0.10
0.10
0.13
0.16
0.16
0.13
0.11
0.11
0.10
0.10
0.10
0.08

0
0.07
0.06
0.07
0.09
0.09
0.08
0.09
0.11
0.09
0.09
0.11
0.09
0.11
0.08
0.12
0.08
0.12
0.11
0.10
0.10
0.08
0.08
0.08
0.09
0.10
0.09
0.09
0.09
0.09
0.09
0.09

conditions, were generated in parallel with ethanol. Furthermore,

the relatively high amounts of 1-butanol produced from the
clostridia species (Tables 2 and 3) indicate that clostridia use the
fermentation pathway for 1-butanol and acetone formation to
ferment carbohydrates to ethanol and to a lesser extend to 1butanol as previously has been suggested [5].
The amounts of ethanol produced by each microbe were quite
different although the initial glucose content of the medium was
the same (200 mg/dL). The observed differences in microbes
ability to produce ethanol and higher alcohols during fermentation

probably reect differences in their metabolic efciency under

anaerobic conditions in respect to the substrate catabolism.
These results indicate that the glucose content is not the only
determinant of the produced ethanol although it should have a
signicant role in the procedure, and the process is potentially
affected by other factors as well. The most obvious factor is the
microbe type even for species becoming of the same genera (herein
C. pergrigens and sporogenes). The patterns of the higher alcohols
have also shown signicant differences between the different
microbes cultures indicating the exibility offered by branched

Fig. 2. (A) Four descriptors (4D) model and; (B) one descriptor (1D) model for ethanol production by C. perfrigens.

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

194

fermentation pathways to microbes, so as, to cover their needs for

energy and redox balance [5].
3.4. Models of microbial ethanol production
3.4.1. Models for C. perfrigens
The initial model for C. perfrigens was built using four
independent variables, namely the 1-propanol, 1-butanol, isobutanol and methyl-butanol concentrations respectively. Then the
less signicant variable (methyl-butanol with descriptor power
1.50) was skipped and a model with three independent variables
(1-propanol, 1-butanol, isobutanol) was built and, so on, isobutanol was skipped (with descriptor power 82.94) resulting to the
creation of the model with two independent variables (1-propanol,
1-butanol with descriptor power 100 and 87.17 respectively). The
signicance of 1-propanol as a descriptor (descriptor power 100) in
this case is so powerful that a considerable satisfactory model
could be created by using it as the only independent variable.
Thereinafter are given for comparison the models by using four
independent variables (Spearman Rank Correlation, r = 0.95)
(Eq. (1)) and the most powered descriptor variable (Spearman
Rank Correlation, r = 0.95) (Eq. (2)). In Fig. 2 the graphs and the
statistical parameters of the relevant models described by the
Eqs. (1) and (2) are showed:
0:30  Isobutanol  0:01  Methyl-butanol 0:03
Ethanol 0:08  1Propanol 0:03  1Butanol

(1)

Ethanol 0:11 0:047  1Propanol

(2)

Thereinafter are given for comparison the models by using four

independent variables (Spearman Rank Correlation, r = 0.55
(Eq. (3)) and the two most powered descriptor variable (Spearman
Rank Correlation, r = 0.39) (Eq. (4)). In Fig. 3 the graphs and the
statistical parameters of the relevant models described by the
Eqs. (3) and (4) are showed:
0:61  Methyl-butanol  0:07  Isobutanol 0:05
Ethanol 0:16  1Propanol  0:07  1Butanol

(3)

Isobutanol 0:09
Ethanol 0:15  1Propanol  0:14 

(4)

3.4.3. Models for E. coli

The initial model for E. coli was built using four independent
variables, 1-propanol, 1-butanol, isobutanol and methyl-butanol
concentrations respectively. Then the less signicant variable
(isobutanol with descriptor power 7.65) was skipped and the
model with three independent variables (1-propanol, 1-butanol,
methyl-butanol) was built and, so on, 1-propanol with descriptor
power 28.14 was skipped resulting to the development of the
model with two independent variables (1-butanol, methyl-butanol
with descriptor power 100 and 53.68 respectively).
Thereinafter are given for comparison the models by using four
independent variables (Spearman Rank Correlation, r = 0.70)
(Eq. (5)) and the two most powered descriptors (Spearman Rank
Correlation, r = 0.56) (Eq. (6)). In Fig. 4 the graphs and the
statistical parameters of the relevant models described by the
Eqs. (5) and (6) are showed:
1:61  1Butanol 1:15  Methyl-butanol 0:15

3.4.2. Models for C. sporogenes

The initial model for C. sporogenes was built using four
independent variables, 1-propanol, 1-butanol, isobutanol and
methyl-butanol concentrations, respectively. Then the less signicant variable (methyl-butanol with descriptor power 21.95) was
skipped and a model with three independent variables (1propanol, 1-butanol, isobutanol) was built and, so on, isobutanol
was skipped (descriptor power 30.32) resulting to the development of the model with two independent variables (1-propanol, 1butanol with descriptor power 100 and 56.01 respectively).

Ethanol 0:07  1Propanol 0:20  Isobutanol

(5)

Methyl-butanol 0:15
Ethanol 2:25  1Butanol 0:98 

(6)

3.4.4. Aspects of ethanol production modeling

Interestingly the signicance of each higher alcohol as a
descriptor in building the relevant model was not in relevance with
its produced amount. The model corresponding to C. perfrigens for
predicting ethanol production (Eq. (1)) was highly correlated to the

Fig. 3. (A) Four descriptors model and; (B) two descriptor model for ethanol production by C. sporogenes.

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

195

Fig. 4. (A) Four descriptors model and; (B) two descriptor model for ethanol production by E. coli.

production of 1-propanol (descriptor signicance 100) making

feasible the use of 1-propanol alone as a quite satisfactory
descriptor for calculating ethanol (Eq. (1)), although it was the
second most abundant produced higher alcohol. However, the
usage of three-descriptors has improved the model and has
introduced 1-butanol (the most abundant) and even isobutanol
(third in abundance) as relatively important determinants. The
decision on which model could be the most applicable depends on
the experimental error for the determination of each descriptor in
agreement with the general consideration that the simplest the
model the easier its application.
The model corresponding to E. coli for predicting ethanol
production (Eq. (5)) was presumably correlated to the production
of 1-butanol (descriptor signicance 100) and of methyl-butanol
(descriptor signicance 53.68). Even by using these two descriptors the model was quite satisfactory (Eq. (6)). The use of three
descriptors (1-butanol, methyl-butanol and 1-propanol) resulted
practically to the best t, while the skip of isobutanol as a
descriptor did not affect the results.
Finally, the model corresponding to C. sporogenes appeared to
be the most complicated one. The signicant deviation of the
experimental values from the relative theoretical ones indicates
that the system did not follow the linear regression but a more
complex one. Meanwhile the model using three descriptors (1propanol, 1-butanol and isobutanol) appeared to be the most
satisfactory.
Our contribution introduces the alcohols 1-butanol, 1-propanol, isobutanol, and methyl-butanol as potential biomarkers for
quantifying microbial ethanol production and not only as
qualitative indicators of the process. This is the main difference
of our study compared to previously reported results by other
groups. The study of Nanikawa and colleagues has correlated the
ratio of 1-propanol to ethanol detected in rat blood and skeletal
muscle after death to the postmortem formed ethanol [13]. Felby
and Nielsen in their study have correlated the blood ethanol and 1propanol levels to the state of putrefaction of the body [14], while
the study of Moriya and Hashimoto has correlated ethanol to 1propanol in respect to the stage of putrefaction of the brain of
drowned persons [15]. Finally 1-butanol has been suggested by
Gubala as an indicator of how long a dead body has been laid in
water [16]. Our results are in agreement with the aforementioned
studies in identifying 1-propanol and 1-butanol as main indicators

of microbial ethanol production. Furthermore, our results identify

the set of alcohols, 1-propanol, 1-butanol and isobutanol, and the
set of alcohols, 1-butanol, 1-propanol and methyl-butanol, as
signicant biomarkers of ethanol production by the clostridial
species (C. perfrigens and C. sporogenes) and E. coli respectively.
3.5. Applicability of the models in human blood and plasma products
The applicability of the constructed models was tested by
performing anaerobic cultures of normal human blood and plasma
inoculated with each one of the bacterial species at 25 8C. The
alcohols content was determined by HS-GCFID at days 0, 1, 2, 3, 5,
7, 11 and 15 and the experimental results were compared to the
theoretical results deriving from the relevant models for C.
pergens, C. sporogenes and E. coli, respectively.
The following blood culture systems were tested in parallel:
I. whole human blood with EDTA inoculated with C. perfrigens;
II. whole human blood with citric ions inoculated with C.
perfrigens;
III. human plasma inoculated with (A) E. coli; (B) C. sporogenes; and
(C) C. perfrigens;
IV. human plasma spiked with glucose so as to achieve 200 mg/dL
initial concentration inoculated with (A) E. coli; and (B) C.
sporogenes.
3.5.1. Alcohols production
In the blood culture system I ethanol and 1-propanol were
produced during the fermentative period, while in the presence of
citric ions (system II) ethanol was produced after the fth day of
fermentation, without production of detectable amounts of any
other alcohol.
When human plasma was used as the culture medium and was
inoculated with E. coli (IIIA) and C. perfrigens (IIIC), ethanol and 1propanol were the only detectable alcohols produced during
fermentation. The system IVA, consisting of human plasma
which was spiked with glucose and inoculated with E. coli,
produced higher amounts of ethanol than the system IIIA
consisting of human plasma inoculated with E. coli without
additional glucose. Interestingly, the ethanol produced in the
system IVA was signicantly lower than the ethanol amount
produced in the E. coli cultures with BHI as the culture medium

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

196

Eq. (1). The experimental values cannot be predicted successfully

by PLS regression and the most appropriate tting model is a 2nd
degree polynomial (R2 = 0.79) implying the difculty in successfully predicting its behavior. An alternative methodology might be
to remove the experimental values from the rst two days,
although our intension was to approach the volatiles relationship
during the whole experimental time, which will result in a linear
relationship with R2 = 0.80.
In the experiment with human plasma inoculated with C.
perfrigens (IIIC) the concentration of produced 1-propanol remains
almost constant from day one to day fteen. The application of the
relevant model (Eq. (1)) gives a maximum error of 22% (average
16%). The model also conrms the linear relationship between
ethanol and 1-propanol, at rst approximation, showing R2 = 0.78
when the high experimental value for ethanol measured on the
rst day was excluded.
The model of E. coli has a quite low mean square error
(indicating the difference between the true values of the
concentrations and the estimated ones) which is suggestive of a
successful prediction. The results have showed that the relevant
model (Eq. (5)) is applicable for human plasma spiked with glucose
(system IVA). The maximum error between the calculated and
theoretical values of ethanol concentration is approximately 25%
(average 22%) for IVA. When the model is applied to the system

(Table 4), although the initial glucose concentration in both culture

media was the same (200 mg/dL). The systems IIIB and IVB
consisting of human plasma inoculated with C. sporogenes did not
produce ethanol or other alcohols during the 15 days of incubation,
irrespectively of the additional glucose (system IVB).
These results emphasize that differences in the composition of
the culture medium (BHI, whole human blood and human plasma)
can affect the amount of the produced ethanol by each microbe and
they are in agreement with our previous observation (Section 3.3).
The cease in the higher alcohols production in C. pefrigens human
blood cultures in the presence of citric ions (system II) may be
attributed to the exibility offered by the amino acids fermentation
pathways (which generate the higher alcohols) which can be omitted
under certain conditions, as suggested earlier [5]. The determination
of 1-propanol as the only detectable higher alcohol produced in the
microbial blood and plasma cultures could be attributed to its ability
to be produced from many different substrates and fermentation
pathways, in response to changing environmental conditions, in
agreement with previous suggestion [5].
3.5.2. Testing of the models
The results of the system consisting of whole human blood
inoculated with C. perfrigens (system I) appear to be very
complex compared to the relevant model as this was described by

Table 5
Results for the calculated ethanol concentrations after applying each model (Eqs. (1)(6)) in postmortem cases with the standard error (E%) for each case.
#

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

Manner of death

Violent
Natural
Natural
Violent
Natural
Natural
Undetermined
Violent
Undetermined
Undetermined
Undetermined
Natural
Natural
Violent
Natural
Undetermined
Natural
Violent
Undetermined
Undetermined
Undetermined
Undetermined
Violent
Undetermined
Undetermined
Undetermined
Undetermined
Violent
Undetermined
Undetermined
Natural
Violent
Undetermined
Undetermined
Undetermined
Undetermined
Undetermined
Natural
Undetermined
Undetermined

Measured ethanol (g/L)

0.10
0.16
0.16
0.19
0.20
0.21
0.21
0.22
0.25
0.29
0.35
0.39
0.42
0.42
0.42
0.47
0.48
0.52
0.58
0.59
0.60
0.62
0.62
0.63
0.64
0.68
0.68
0.69
0.69
0.71
0.72
0.73
0.75
0.81
0.87
0.91
0.94
0.98
1.08
2.58

C.
perfrigens,
Eq. (1)

C. perfrigens,
Eq. (2)

E. coli, Eq. (5)

E. coli, Eq. (6)

C.
sporogenes,
Eq. (3)

C. sporogenes,
Eq. (4)

Ethanol
(g/L)

E%

Ethanol
(g/L)

E%

Ethanol
(g/L)

E%

Ethanol
(g/L)

E%

Ethanol
(g/L)

E%

Ethanol
(g/L)

E%

0.13
0.15
0.09
0.22
0.06
0.30
0.20
0.27
0.14
0.29
0.36
0.33
0.44
0.54
0.40
0.84
0.72
0.65
0.85
0.29
0.63
1.12
1.02
1.01
0.43
0.43
0.41
0.42
0.25
0.29
0.46
0.67
1.00
0.30
0.95
0.92
1.13
0.43
0.29
1.08

26
8
43
18
70
45
6
23
43
1
2
16
5
28
6
79
50
26
47
52
6
80
64
60
34
36
39
39
64
60
35
8
34
64
9
2
20
56
73
58

0.05
0.05
0.07
0.00
0.09
0.05
0.03
0.03
0.04
0.04
0.06
0.03
0.06
0.19
0.11
0.32
0.29
0.25
0.28
0.04
0.23
0.20
0.45
0.42
0.05
0.11
0.11
0.08
0.02
0.04
0.14
0.26
0.46
0.03
0.42
0.42
0.33
0.10
0.04
0.49

151
129
145
99
146
78
116
87
117
85
82
91
85
56
75
32
39
51
52
93
61
67
27
33
91
84
84
89
98
94
81
64
38
96
52
54
65
90
96
81

0.24
0.71
0.20
0.32
0.18
0.39
0.29
0.36
0.25
0.38
0.43
0.61
0.49
0.59
0.66
0.84
0.75
0.69
0.83
0.37
0.67
0.98
1.01
0.99
1.22
0.50
1.11
0.48
0.34
0.37
1.03
0.71
1.00
0.74
1.56
0.93
1.72
0.49
0.37
1.06

128
344
26
69
12
85
38
64
1
32
23
57
16
41
55
79
57
34
44
37
13
59
63
58
91
27
63
31
51
47
43
3
33
9
80
3
83
50
65
59

0.15
0.80
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.45
0.15
0.15
0.33
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.89
0.15
1.03
0.15
0.15
0.15
0.82
0.15
0.15
0.63
0.99
0.15
0.94
0.15
0.15
0.15

44
399
8
20
25
29
29
31
41
48
57
16
64
64
21
68
69
71
74
75
75
76
76
76
40
78
51
78
78
79
13
79
80
22
13
83
0
85
86
94

0.24
0.24
0.18
0.43
0.11
0.58
0.30
0.52
0.28
0.57
0.63
0.52
0.61
1.05
0.85
1.49
1.42
1.28
1.33
0.55
1.21
0.98
1.95
1.85
0.87
0.80
0.76
0.67
0.48
0.56
0.92
1.30
2.00
0.56
1.84
1.84
1.61
0.74
0.56
2.08

131
53
8
128
45
176
41
139
11
97
81
34
45
151
102
217
196
149
129
6
104
58
214
193
35
17
12
2
31
21
28
78
166
32
112
103
71
24
49
19

0.27
0.29
0.21
0.44
0.15
0.59
0.31
0.53
0.31
0.57
0.63
0.53
0.59
1.03
0.78
1.42
1.37
1.25
1.25
0.56
1.18
0.83
1.86
1.76
0.56
0.78
0.78
0.66
0.49
0.57
0.88
1.26
1.91
0.54
1.76
1.76
1.32
0.73
0.56
1.99

156
82
28
136
27
179
48
144
21
100
80
36
39
145
84
202
186
142
115
5
97
33
200
179
13
15
15
5
30
20
23
73
155
33
103
95
40
26
48
23

The original ethanol concentrations measured for each case are also provided along with the manner of death.

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

of human plasma (IIIA) the model loses its power probably due to
the fact that ethanol and 1-propanol values are strongly correlated
exhibiting a steady proportion of 1:10.
Our results indicate that the constructed models have the
potential to be applied when different culture media are used,
under controlled experimental conditions. Factors such as the type
of microbial species, the glucose content and the medium
composition apparently affect the procedure of microbial ethanol,
1-butanol and higher alcohols production.
3.6. Use of the models in real postmortem cases
We retrospectively looked our chromatograms archives of the
last six years and we selected 40 chromatograms corresponding to
postmortem cases, having presence of other alcohols during the
original forensic ethanol analysis. The classication of these cases
as resulted from the retrospective review of the forensic pathologists
reports was: 10 cases of natural deaths; 8 cases of violent deaths;
and 22 cases of undetermined cause of death. Furthermore, 39 cases
had marked putrefactive phenomena at autopsy while the other one
had extensive traumatic lesions (Case no 4) making easier the
invasion of microbes. We calculated for these cases the concentrations of higher alcohols and 1-butanol. In Table 5 the original ethanol
concentrations measured for each case and the calculated ethanol
concentrations after applying each model (Eqs. (1)(6)) are provided
along with the standard error for each case.
The models corresponding to C. perfrigens estimated the
microbial produced ethanol with a standard error <40% for 27
out of the 40 cases (68%), 26 of them with marked putrefaction (96%).
Additionally, 21 of these cases (78%) had ethanol lower than 0.7 g/L.
The models corresponding to E. coli estimated the microbial
produced ethanol with a standard error <40% for 25 out of the 40
cases (63%), 24 of them with marked putrefaction (96%). Moreover,
18 of these cases (72%) had ethanol lower than 0.7 g/L.
The models corresponding to C. sporogenes estimated the
microbial produced ethanol with an error <40% for 18 out of the 40
cases (45%) all presented with putrefaction at autopsy, while 12 of
these cases (67%) had ethanol lower than 0.7 g/L.
It is worth mentioning that 28 out of the 29 cases (97%) having
original ethanol concentration lower than 0.7 g/L succeeded a
standard error <40% in predicting the microbially produced
ethanol by at least one of the models.
The cases were selected due to the co-detection of signicant
amounts of volatiles along with ethanol during the original
ethanol analysis a factor making the relevant cases suspicious
for microbial ethanol production, as it was suggested [24]. Then
it was proved that the majority of the selected cases had marked
putrefaction at autopsy another aggravating factor for
microbial activity and ethanol neo-formation [14,15,24,25].
Our results have shown that the models could effectively be
applied in postmortem cases especially when marked putrefaction is present. In addition, the models of C. perfrigens show
better applicability than these of E. coli and than those by C.
sporogenes. Yet, the models are applied more satisfactory when
ethanol concentrations are lower than 0.7 g/L. It is generally
accepted that ethanol concentrations lower than 0.7 g/L is more
possible to be of microbial origin for the majority of cases with
neo-formation of ethanol [2,3].
However, it has to be underlined that so far the study of real
postmortem cases meets serious limitations. As the ethanol
concentrations increase, it is possible part of the detected ethanol
to be of ante mortem origin making the proper interpretation of
postmortem ethanol analysis difcult. In such a case it is extremely
challenging and premature to suggest that one model is more
suitable than another. On the other hand, other microbes might
generate different alcohols patterns resulting to different models.

197

This might to explain the major discrepancies observed between

measured and calculated ethanol concentrations for some of the
presented cases.
Ideally, it should be known which microbial species have
been present or have been activated in each post-mortem case.
Practically, more mathematical models, covering a large
spectrum of bacterial or fungal species should be available.
Moreover, the identication of more inuencing factors and
the quantication of their impact on the procedure would
eventually
result
to
the
construction
of
more
accurate models which could be applied more efciently to
real cases.
Under these views it is naive to suggest that such a complicated
process, with so many potential effectors, could be entirely
covered by the simple models resulting from the study of only
three microbes. However, this study represents only an initial
approach, albeit with some useful outcomes; it is promising that
even such a simplied approach, highly challenging at the present,
might provide a methodical insight into this huge, long lasting
problem.
4. Concluding remarks
To our knowledge the present contribution is a rst approach to
the quantication of microbial ethanol production by demonstrating a mathematical model of the procedure in cases where other
alcohols are produced simultaneously with ethanol. The microbial
biosynthesis of ethanol in parallel with the biosynthesis of 1butanol and the higher alcohols (1-propanol, isobutanol, 2-methyl1-butanol and 3-methyl-2-butanol) identies these alcohols as
biochemical biomarkers suitable for quantifying the microbial
ethanol production.
The correlation between the levels of the alcohols and the
amount of the microbially produced ethanol is expressed by
mathematical models (equations).
The alcohols 1-propanol and 1-butanol appear to have a
preponderant role as descriptors of the relevant models.
The most obvious factor affecting the produced ethanol, even
for species within the same genus (such as C. perfrigens and C.
sporogenes), is the microbe type.
Differences in the composition of the culture medium (despite
the initial glucose concentration) result in the production of
different amounts of ethanol and other alcohols from the same
species.
The overall conclusion of the reported experimental study is
that mathematical modeling of the microbial ethanol production
is feasible, at least partially. Nevertheless, given the complexity
of the process, extrapolation of the results to everyday practice
requires
ongoing
work
on
the
identication
and
quantication of the factors inuencing the postmortem ethanol
production.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.forsciint.2011.03.003.
References
[1] C.L. ONeal, A. Poklis, Postmortem production of ethanol and factors that inuence
interpretation, Am. J. Forensic Med. Pathol. 17 (1996) 820.
[2] K. Ziavrou, V.A. Boumba, T. Vougiouklakis, Insights into the origin of postmortem
ethanol, Int. J. Toxicol. 24 (2005) 6977.
[3] F.C. Kugelberg, A.W. Jones, Interpreting results of ethanol analysis in postmortem
specimens: a review of the literature, Forensic Sci. Int. 165 (2007) 1029.

198

V.A. Boumba et al. / Forensic Science International 215 (2012) 189198

[4] J.E.L. Corry, Possible sources of ethanol ante- and postmortem: its relation to the
biochemistry and microbiology of decomposition, J. Appl. Bacteriol. 44 (1978) 156.
[5] V.A. Boumba, K. Ziavrou, T. Vougiouklakis, Biochemical pathways generating
post-mortem volatile compounds co-detected during forensic ethanol analysis,
Forensic Sci. Int. 174 (2008) 133151.
[6] D. Yamima, H. Motani, K. Kamei, Y. Sato, M. Hayakawa, H. Iwase, Ethanol
production by Candida albicans in post-mortem human blood samples: effect
of blood glucose level and dilution, Forensic Sci. Int. 164 (2006) 116121.
[7] J. Chang, S.E. Kollman, The effect of temperature on the formation of ethanol by
Candida albicans in blood, J. Forensic Sci. 34 (1989) 105109.
[8] J.J. Saady, A. Poklis, H.P. Dalton, Production of urinary ethanol after sample
collection, J. Forensic Sci. 38 (1993) 14671471.
[9] H.A. Sulkowski, A.H. Wu, Y.S. McCarter, In-vitro production of ethanol in urine by
fermentation, J. Forensic Sci. 40 (1995) 990993.
[10] B.M. Appenzeller, M. Schuman, R. Wennig, Was a child poisoned by ethanol?
Discrimination between ante-mortem consumption and post-mortem formation,
Int. J. Legal Med. 122 (2008) 429434.
[11] A.C. Gruszecki, C.A. Robinson, S. Kloda, R.M. Brissie, High urine ethanol and
negative blood and vitreous ethanol in a diabetic woman: a case report, retrospective case survey, and review of the literature, Am. J. Forensic Med. Pathol. 26
(2005) 9698.
[12] G.D. Amick, K.H. Habben, Inhibition of ethanol production by Saccharomyces
cerevisiae in human blood by sodium uoride, J. Forensic Sci. 42 (1997) 690692.
[13] R. Nanikawa, K. Ameno, Y. Hashimoto, K. Hamada, Medicolegal studies on alcohol
detected in dead bodies alcohol levels in skeletal muscle, Forensic Sci. Int. 20
(1982) 133140.
[14] S. Felby, E. Nielsen, Congener production in blood samples during preparation and
storage, Blutalkohol 32 (1995) 558.

[15] F. Moriya, Y. Hashimoto, Postmortem production of ethanol and n-propanol in the

brain of drowned persons, Am. J. Forensic Med. Pathol. 25 (2004)
131133.
[16] W. Gubala, n-Butanol in blood as the indicator of how long a dead body lay in
water, Forensic Sci. Int. 46 (1990) 127128.
[17] C.L. Morris-Kukoski, E. Jagerdeo, J.E. Schaff, M.A. LeBeau, Ethanol analysis from
biological samples by dual rail robotic autosampler, J. Chromatogr. B Analyt.
Technol. Biomed. Life Sci. 850 (2007) 230235.
[18] D. Zuba, W. Piekoszewski, J. Pach, L. Winnik, A. Parczewski, Concentration of
ethanol and other volatile compounds in the blood of acutely poisoned alcoholics,
Alcohol 26 (2002) 1722.
[19] L. Kristoffersen, L.E. Stormyhr, A. Smith-Kielland, Headspace gas chromatographic
determination of ethanol: the use of factorial design to study effects of blood
storage and headspace conditions on ethanol stability and acetaldehyde formation in whole blood and plasma, Forensic Sci. Int. 161 (2006) 1511577.
[20] .K.J. Ryan, C.G. Ray (Eds.), Sherris Medical Microbiology, 4th ed., McGraw Hill, 2004
[21] A.W. Jones, L. Hylen, E. Svensson, A. Helander, Storage of specimens at 4 degrees C
or addition of sodium uoride (1%) prevents formation of ethanol in urine
inoculated with Candida albicans, J. Anal. Toxicol. 23 (1999) 333336.
[22] E.R. Kuhn, A.K. Vickers, S.E. Ebleler, D.M. Ahlgren, J.H. Thorngate, Complete
separation and quantitation of fusel oils by capillary GC, Application, Agilent
Technologies, Inc. 2003.
[23] B.S. de Martinis, C.C.S. Martin, Automated headspace solid-phase microextraction
and capillary gas chromatography analysis of ethanol in post-mortem specimens,
Forensic Sci. Int. 128 (2002) 115119.
[24] G. Scopp, Postmortem toxicology, Forensic Sci. Med. Pathol. 2 (2010) 314325.
[25] D.M. Butzbach, The inuence of putrefaction and sample storage on post-mortem
toxicology results, Forensic Sci. Med. Pathol. 6 (2010) 3545.