Can a Postmortem Ethanol Result Be Falsely Low? Microbial Oxidation and an HS-GC–MS Study

Short answer: Yes, ethanol in a postmortem specimen can be lost after death or collection under particular microbial or enzymatic conditions. But a suspiciously low result, a negative result, or an unusual internal-standard signal cannot by itself establish that any particular deceased person’s original blood ethanol concentration was higher. Edstam, Sundqvist, and Kronstrand’s headspace gas chromatography–mass spectrometry (HS-GC–MS) study provides a concrete analytical warning: approximately 2% of authentic samples exhibited abnormal responses from isotope-labeled internal standards, and a selected experimental postmortem sample oxidized added ethanol during refrigerated storage. These findings justify case-specific review of specimen integrity, laboratory controls, and corroborating matrices—not automatic reconstruction of a missing blood alcohol concentration.

The primary source is Edstam, Sundqvist, and Kronstrand, “Quantitation of alcohols and acetone in postmortem blood and urine using headspace gas chromatography mass spectrometry,” Journal of Analytical Toxicology, DOI: 10.1093/jat/bkaf076 (published online in 2025; journal volume 50, 2026). This article examines its directly observed results and explains what they do—and do not—support in a death investigation.

What was actually tested?

The authors developed and validated a laboratory method for postmortem blood and urine. The method directly quantified six volatile compounds: ethanol, methanol, 2-propanol (isopropanol), 1-propanol, 1-butanol, and acetone. This was an analytical-method study with authentic autopsy specimens, quality-control material, and targeted stability experiments. It was not a controlled drinking experiment in living volunteers; no driving tests, observed intoxication ratings, or pre-death serial blood samples were collected for individual decedents.

For routine quantitation, the laboratory dispensed 100 µL of sample with 1,000 µL of internal-standard solution into a 20-mL headspace vial. A gas chromatograph separated volatile chemicals; a mass spectrometer monitored characteristic ions and ion ratios. Selected-ion monitoring provided concentrations for the six validated analytes, while full-scan mass spectra helped identify other volatile compounds. Concentrations reported in mg/mL refer to the tested blood or urine matrix, not a breath-alcohol concentration or a calculated whole-person alcohol dose.

Why do deuterated internal standards matter in forensic ethanol analysis?

An internal standard is added to a specimen so that an analyte’s instrument response can be assessed relative to a known compound. The method employed isotopically labeled, or deuterated, ethanol, acetone, 2-propanol, and 1-butanol as internal standards. The labeled compound behaves similarly to the target under many extraction and instrumental conditions but can be distinguished by its mass spectrum.

This design addresses a special problem with postmortem samples: some older methods use unlabeled 1-propanol as an internal standard, even though 1-propanol itself can be present in decomposing samples. A compound that exists naturally in the specimen is a problematic reference unless that interference is recognized and controlled. Isotope dilution also helps correct certain matrix and sample-handling variations, although it does not make a specimen biologically stable or eliminate every source of bias.

There is an important qualification. If ethanol-d₆, the labeled reference, is consumed in a compromised specimen, the analyte-to-reference signal ratio may become misleading. That is a quality-control warning, not a calibrated formula for recovering the decedent’s earlier ethanol concentration. The direction and magnitude of any reporting error depend on how the specimen, endogenous ethanol, internal standard, instrument, and software responded.

What did the authors find about microbial or enzymatic ethanol loss?

During routine work, some authentic postmortem specimens showed unexpectedly low responses for ethanol-d₆ and 2-propanol-d₆, sometimes accompanied by increased acetone-d₆. The authors reported this pattern in approximately 2% of authentic samples. They investigated it using the alcohol-dehydrogenase inhibitor 4-methylpyrazole (4-MP) and also considered catalase inhibition. Their experiments supported enzymatic conversion of ethanol-d₆ and of 2-propanol-d₆ to acetone-d₆ during sample processing under the conditions studied.

The researchers also selected a postmortem blood sample identified as having microorganisms present, fortified it with ethanol, 2-propanol, and acetone, and compared measurements at preparation with results after 14 days of refrigerated storage. They reported that all added ethanol had been oxidized in that experimental sample, with acetone increasing as 2-propanol was oxidized. It is strong evidence of what can happen in a susceptible specimen; it is not evidence that all refrigerated autopsy samples behave this way, nor that 2% of deaths have a falsely negative blood alcohol result.

Three quantities must not be conflated: the fraction of samples with atypical internal-standard responses; the size of ethanol loss in a selected experimental specimen; and an individual’s unknown antemortem ethanol concentration. The paper establishes the first two under stated conditions. It does not supply an equation that identifies the third.

What did the validation and method comparison establish?

The investigators validated the analytical measurement rather than a clinical or legal conclusion about intoxication. Their calibration model was quadratic and weighted by inverse concentration. Precision experiments used water, donor urine, and bovine blood as well as an additional authentic postmortem case. The authors expressly noted deviations from parts of the referenced validation-standard design, including five calibration levels rather than six and the use of bovine rather than authentic postmortem blood for some tests.

These figures are method- and control-level-specific. A stated 2.8% expanded analytical uncertainty is not the uncertainty in a death investigation’s time of drinking, whether microbial formation occurred, or whether blood ethanol at death equaled ethanol at a prior incident. It also does not bound preanalytical degradation that occurred before the laboratory measurement.

The laboratory compared 152 blood and 139 urine autopsy specimens using its HS-GC–MS method and an existing dual-column HS-GC–FID procedure. Among 99 ethanol-positive blood comparisons, 72 differed by less than ±5% between methods; among 79 ethanol-positive urine comparisons, 65 met that benchmark. No blood or urine sample differed by more than 20% in the comparison reported. Mean differences were approximately −0.96% for blood and −2.8% for urine under the authors’ subtraction convention.

The expanded ethanol uncertainty estimate was 2.8% for HS-GC–MS at a 0.20 mg/mL control level and 4.6% for the older HS-GC–FID method at a 0.50 mg/mL control level. Because the control concentrations and methods were not identical, this does not prove every GC–MS installation will be more accurate than every GC–FID installation. It does show that identification chemistry, the internal standard, the matrix, and the actual laboratory’s validation records matter.

Can 1-propanol or 1-butanol prove that detected ethanol was made after death?

No. Both are potentially useful indicators of microbial processes, but co-detection is not source attribution for each ethanol molecule. In the first six months of routine operation, the investigators described 2,500 cases. Of the femoral-blood specimens, 735 were ethanol-positive. Within that ethanol-positive set, approximately 22% contained quantifiable 1-propanol and approximately 10% contained quantifiable 1-butanol. Those figures describe co-occurrence in this particular laboratory population and reporting range; they do not give the fraction of all cases in which ethanol was formed after death.

Microbial metabolism depends on microorganisms present, substrates, temperature, storage conditions, time, and the state of the remains. Different bacteria and fungi can produce different volatile profiles. Therefore, a low 1-propanol result is not universal proof of putrefaction, and an absent 1-propanol result is not universal proof that all ethanol was consumed before death.

These are reported counts of positive specimens, not mutually exclusive case categories or a prevalence estimate that can be transferred to other jurisdictions. The routine table does not independently establish the temporal sequence of drinking, death, sampling, or microbial growth.

How does a low ethanol result differ from evidence of ethanol formation?

Two seemingly opposite processes can complicate postmortem toxicology. Formation means microbes or other biological activity generated ethanol in the remains or collected material after death. Degradation means ethanol already present was consumed or altered. They may occur under different conditions, and a sample’s ultimate reported concentration is not necessarily a reliable record of its earlier concentration.

The Edstam study specifically exposed a degradation mechanism through unusual deuterated-reference signals and inhibitor experiments. It also improved routine quantitation of potential putrefactive alcohols. Neither observation alone determines whether a given person drank before death. The authors proposed that an independent biomarker such as ethyl sulfate could help investigate the origin of ethanol in selected cases; ethyl sulfate was not one of this study’s six routinely quantified volatile analytes. A separate validated biomarker result must not be implied when none was obtained.

A death investigation should therefore ask both whether ethanol could have been generated and whether a portion of ethanol could have been lost. Those are questions about sample history and biological interpretation, beyond the chromatography’s demonstration that a particular molecule was present at analysis.

Could ordinary headspace heating alone explain major ethanol loss?

The investigators compared water and postmortem blood fortified at 2.0 mg/mL during headspace incubation at 70°C for 7, 15, and 25 minutes. Across the conditions studied, mean differences remained below 0.5% from the nominal value, though there was a small decreasing trend over time. This experiment helps separate routine instrument equilibration from the pronounced, specimen-specific biological activity observed elsewhere in the study.

The point is not that heat is categorically irrelevant, but that the tested incubation conditions did not reproduce large losses across these prepared controls. Evidence of microbial action in a particular case needs its own laboratory and specimen documentation. It would be incorrect to borrow the 14-day selected-specimen result as a routine correction for every heated headspace vial.

What documentation matters when a postmortem blood alcohol result is disputed?

The decisive materials are often more specific than the final reported number. A reviewer should identify the anatomical blood-collection site, the exact matrix, postmortem interval, collection date and time, preservative and tube details, refrigeration and shipping history, decomposition observations, and whether a second matrix was available. Postmortem toxicology requires a specimen-specific analysis rather than a generic rule that every blood measurement is interchangeable.

For this method in particular, the reviewer should request instrument chromatograms, selected-ion traces, qualifier-ion ratios, calibration, quality-control results, internal-standard peak areas, batch annotations, and any reanalysis or inhibitor experiment. A sudden depression of ethanol-d₆ or 2-propanol-d₆ deserves a reasoned explanation. It is not sufficient to assume that any laboratory’s standard chromatographic signal passed without examining the relevant run. For a background framework, see blood alcohol testing by headspace chromatography.

A positive urine ethanol concentration must also be kept distinct from a contemporaneous femoral blood concentration; the fluids have different kinetics and histories. Cross-matrix comparison may expose an inconsistency worth investigating, but it is not a validated equation for back-calculating a whole-blood BAC at the time of a crash, assault, or death. Sound specimen collection and custody documentation is integral to the scientific opinion.

What can the method establish—and what remains a forensic inference?

Direct measurement: Under the laboratory’s validated conditions, the method identifies and quantifies selected volatile analytes in the submitted postmortem blood or urine specimen, with analyte-specific analytical uncertainty. Quality evidence: Abnormal labeled-standard behavior can suggest a matrix-dependent or enzymatic problem that requires investigation. Biological inference: Interpreting decomposition, postmortem formation, or depletion requires additional evidence about the specimen and its history.

Not established: This study does not measure a living person’s breath alcohol; it does not assign a blood ethanol concentration at an earlier scene; it does not validate an individual time-since-drinking estimate; it does not score subjective intoxication, gait, divided attention, reaction time, or driving behavior; and it does not determine an individual’s manner or cause of death. A numerical method-validation uncertainty does not encompass all of those separate inferential questions.

In litigation, the scientifically defensible question is whether the documented result and the relevant sample history support the proposed conclusion, not whether one published experiment creates a universal basis to increase or dismiss a decedent’s measured ethanol concentration. A witness should separate the analytical observation, possible degradation pathway, corroborative findings, and uncertainty in the case-specific opinion.

Frequently asked forensic questions

Can a postmortem blood sample become ethanol-negative after initially containing ethanol?

It is biologically possible under some conditions. In one selected fortified postmortem sample, added ethanol was entirely oxidized over 14 refrigerated days. That experiment does not demonstrate the starting concentration or history of a different decedent’s sample.

Does a low ethanol-d₆ signal prove the decedent was intoxicated?

No. It may flag a compromised internal-standard response or possible enzymatic activity. It provides no independent evidence that the deceased person consumed alcohol, at any specific dose, before death.

Does detecting 1-propanol establish that all ethanol was produced by decomposition?

No. A higher-alcohol finding can prompt evaluation for microbial activity, but the mixture does not quantify the proportion of ethanol produced after death. Other matrices and markers may be necessary.

Can a negative 1-propanol result prove that the ethanol was antemortem?

No. Absence at the method’s reporting threshold does not exclude microbial processes, different substrates, different organisms, or timing-dependent changes.

Is HS-GC–MS automatically superior to HS-GC–FID in every laboratory?

No. The paper documents an identification advantage and particular validated performance for its isotope-dilution mass-spectrometric procedure. Actual precision, uncertainty, interferences, controls, and matrices must be compared at relevant concentrations.

Can the 2.8% analytical uncertainty be applied to any decedent’s earlier BAC?

No. The percentage is a method-specific expanded uncertainty estimate at the stated control level. It does not encompass unmeasured degradation, postmortem production, time-of-collection uncertainty, or reconstruction of blood ethanol at a prior event.

Does this study determine impairment or cause of death from an ethanol result?

No. The measured endpoints were analytical. Individual impairment and causation require independent clinical, behavioral, pathological, scene, and chronological evidence; the study supplies neither an individual impairment threshold nor a causal attribution rule.

Primary research and evidentiary bottom line

Primary study: Edstam M, Sundqvist G, Kronstrand R. Quantitation of alcohols and acetone in postmortem blood and urine using headspace gas chromatography mass spectrometry. Journal of Analytical Toxicology. Online 2025; volume 50 (2026), article bkaf076. DOI: 10.1093/jat/bkaf076.

The useful finding is not a new universal BAC adjustment. It is that a validated postmortem volatiles method can reveal both analytically meaningful co-analytes and a previously underappreciated specimen-specific pathway for internal-standard and ethanol degradation. That insight strengthens the need for transparent specimen handling, chromatographic QA, and carefully bounded testimony. A direct chemical measurement, an analytical warning sign, a reconstruction of pre-death concentration, and an opinion about impairment or death causation remain distinct propositions.