A fermentation experiment using blood-bank material cannot automatically establish the maximum ethanol that could form in an ordinary forensic specimen. Chang and Kollman’s 1989 study found ethanol formation after inoculation with Candida albicans, but the blood contained added dextrose. The authors specifically acknowledged that this prevented assessment of their proposed glucose-based maximum yield.
What was in the experimental blood?
The researchers obtained four blood-bank collection bags of 450-mL capacity. Each contained 2.0 g dextrose, along with citrate, citric acid, phosphate and adenine. The blood was pooled; half was inoculated with C. albicans ATCC 14056 to approximately 10,000 organisms/mL.
They divided the material among 112 tubes of 10-mL capacity. Half contained 100 mg sodium fluoride and 20 mg potassium oxalate; half had no tube additives. Four conditions combined inoculated or uninoculated blood with preserved or unpreserved tubes. Filling occurred under nonsterile conditions, with uninoculated specimens serving as controls.
These were prepared aliquots from pooled material, not 112 people, 112 autopsies or a survey of naturally contaminated forensic specimens. No postmortem anatomical collection site or body decomposition interval applied to this laboratory design.
The proposed ceiling did not survive the substrate check
The discussion calculated an approximate ethanol yield of 0.05% w/v from a starting glucose concentration of 95 mg/100 mL, assuming complete conversion of one mole of glucose to two moles of ethanol. The experiment produced ethanol exceeding that estimate.
The authors then recognized the blood-bank addition of 2.0 g glucose per unit and stated that their maximum-yield hypothesis could not be assessed. The appropriate conclusion is not that fermentation violated the proposed chemical stoichiometry. The assumed substrate concentration did not describe the experimental material.
Neither the calculated 0.05% w/v nor an observed plateau near 0.08% w/v is therefore a demonstrated universal ceiling for postmortem ethanol. A yield depends on the material and conditions tested; the paper did not validate a fixed allowance to subtract from human autopsy blood.
Preservation and temperature still produced informative findings
At 37°C after about 69 hours, the four inoculated, unpreserved specimens contained 0.019–0.053% w/v ethanol. The preserved specimens, whether inoculated or not, had no detectable ethanol at that time. At 22°C, inoculated specimens developed ethanol after several days even with sodium fluoride. A short warm experiment therefore did not predict the longer room-temperature outcome.
At 6°C, no fermentation was detected during the first 35 days. After 182 days, two of four inoculated, unpreserved specimens contained 0.008 and 0.015% w/v, and one uninoculated, unpreserved specimen contained 0.004% w/v. Preserved specimens remained without detected ethanol at 182 days.
Those negative findings are conditional on the experiment and detection threshold. They do not establish that preservatives or refrigeration guarantee stability for every microbial strain, specimen and storage period.
The analytical units and detection threshold matter
Testing used direct-injection gas chromatography with flame-ionization detection, not headspace GC–MS. Methyl ethyl ketone was the internal standard. Calibration used ethanol standards of 0.122, 0.220 and 0.340% w/v, with 0.177% w/v controls at the beginning, middle and end of runs. The minimum detectable ethanol concentration was 0.003% w/v.
The reported concentrations were percentages by weight per volume: 0.05% w/v corresponds to 0.5 g/L. That conversion does not convert a detection threshold into a quantitation limit; the paper did not provide a separate fully characterized numerical quantitation limit or modern validation dataset for every condition.
What can the experiment support in case review?
It supports asking about substrate composition, organism, preservative, temperature and duration together. It demonstrates formation after experimental contamination and preserves important negative control findings. It does not measure how often forensic specimens are contaminated, establish an individual’s alcohol consumption or assign a cause of death.
A complementary human case concerns ethanol increasing during serial testing of preserved postmortem blood. The designs should remain separate.
Primary source
Chang J, Kollman SE. The Effect of Temperature on the Formation of Ethanol by Candida Albicans in Blood. Journal of Forensic Sciences. 1989;34(1):105–109. doi:10.1520/JFS12616J.
Leave a Reply