EVIDENCE LIBRARY
Can Low Blood Alcohol in a Drowning Be Postmortem Fermentation? A 967-Case Study
A 967-case drowning study found four low-positive blood ethanol results with negative urine. Learn what postmortem fermentation evidence can—and cannot—prove.
Short answer: A low ethanol concentration measured in postmortem blood after a drowning does not by itself prove that the person drank alcohol—or that the ethanol was produced after death. In a Finnish study of 967 unintentional drowning deaths, only four cases had a positive blood ethanol result but no detectable ethanol in the available urine sample, a pattern the investigators considered potentially consistent with postmortem production. That was 0.4% of all selected drownings, but 14.3% of the 28 cases with a positive blood ethanol result below 50 mg/dL. Importantly, none of those four cases had an available vitreous ethanol measurement, so the proposed origin was not independently established.
This is an epidemiological finding about how to classify results, not a test that can identify microbial fermentation, reconstruct an individual’s alcohol exposure, establish impairment, or determine why a drowning occurred. The original study is Pajunen, Vuori and Lunetta (2018), Injury Epidemiology, doi:10.1186/s40621-018-0169-4.
What forensic question did the 967-case study ask?
Alcohol-positive drowning statistics can be distorted in two opposite directions. Counting every measurable postmortem blood ethanol concentration as evidence of premortem drinking could overestimate the number of alcohol-exposed victims if microbes formed ethanol after death. Automatically throwing out low concentrations or cases recovered after a set interval could underestimate genuine premortem alcohol exposure.
Pajunen and colleagues examined whether simultaneous interpretation of the available postmortem blood, urine and vitreous-humor ethanol results would identify many possible cases of endogenous ethanol formation. Their primary question was the reliability of population-level estimates of alcohol-positive drowning, not whether a specific victim was incapacitated or how alcohol contributed to an individual fatal event.
The distinction matters in litigation: a reliable laboratory identification of ethanol and a reliable attribution of its origin are separate scientific propositions. Even a technically sound analytical measurement does not alone establish the source of the molecule.
How were the Finnish drowning cases selected and tested?
The investigators conducted a retrospective examination of Finnish medico-legal records from 2000 through 2013. The final analytic set contained 967 unintentional drowning deaths with a blood alcohol concentration (BAC) and an alcohol result from urine, vitreous humor, or both. Not every decedent had all three matrices available.
Forensic specimens were generally taken from femoral blood, urinary bladder contents and vitreous humor at autopsy, preserved using sodium fluoride, and analyzed by dual-column headspace gas chromatography at the same forensic toxicology laboratory. The laboratory used four independent measurements to produce the final mean concentration. Those measurements established ethanol in a specified specimen; they did not measure whether ethanol had been ingested, synthesized by microbes, or produced in a particular quantity after death.
The study relied on recorded concentrations and case details, including estimated time submerged and season. It did not experimentally incubate each specimen, culture microorganisms in every case, or obtain independently confirmed drinking histories for all victims. It was a retrospective observational analysis rather than a controlled experiment in human ethanol production.
What did the blood alcohol results actually show?
Among the 967 included deaths, 623 had some measured blood ethanol, including 595 at or above 50 mg/dL. The remaining 28 positive results were below 50 mg/dL. The following figures are counts in this selected forensic-autopsy cohort, not population-wide drowning incidence or an estimate of the fraction of cases caused by ethanol.
| Measured result or study classification | Cases | Correct denominator and interpretation |
|---|---|---|
| Blood ethanol not detected (BAC = 0 mg/dL) | 344 | 35.6% of 967 selected drownings |
| Any positive blood ethanol (BAC > 0 mg/dL) | 623 | 64.4% of 967; a concentration result, not proof of impairment |
| Blood ethanol at least 50 mg/dL | 595 | 61.5% of 967; 50 mg/dL equals 0.05 g/dL in blood |
| Blood ethanol at least 100 mg/dL | 567 | 58.6% of 967; a subset of the preceding 595 |
| Low-positive blood ethanol: >0 but <50 mg/dL | 28 | 2.9% of 967; includes four urine-negative cases |
| Blood positive, available urine negative: possible postmortem ethanol | 4 | 0.4% of 967; 14.3% of the 28 low-positive cases; origin not proven |
Denominators are consequential. Four of 967 is approximately 0.41%; four of 28 is approximately 14.29%; and four of the 623 blood-positive cases is approximately 0.64%. These are three different questions. A reader should not describe 0.4% as the proportion of blood-positive cases; its denominator here is all 967 included drownings. None of the fractions is a validated general false-positive rate for postmortem ethanol testing.
Why were four low-positive cases considered possible fermentation?
In four decedents, ethanol was measured in blood at 25, 36, 46 or 48 mg/dL while the available urine result was negative. This discordance may be a warning sign: postmortem blood can provide conditions for microbial ethanol production, while urine and vitreous humor may be less susceptible in many circumstances.
But a blood-positive/urine-negative pattern has another plausible explanation. If a living person drinks alcohol shortly before death, blood ethanol may rise before sufficient ethanol enters urine. That is especially important during early absorption. The study itself highlighted a cold-water case with a 46 mg/dL blood result and submersion shorter than one day as an example in which recently ingested alcohol could not be excluded.
Vitreous ethanol was unavailable in all four discordant cases. The researchers therefore did not establish that those concentrations were formed by microbes. Neither a positive blood result nor a negative urine result measures microbial production directly. In an individual case, evidence of specimen handling, decomposition, timing, recent intake, analytical quality, and other matrices must be assessed together.
How should blood, urine and vitreous ethanol be interpreted differently?
These are separate biological matrices with different physiology and postmortem susceptibility. They must not be substituted for one another or treated as three interchangeable measurements of a single legal “BAC.” Blood ethanol reflects the tested blood specimen; urinary ethanol reflects ethanol in a collected bladder specimen influenced by urine formation and timing; vitreous ethanol is a measurement in ocular fluid with its own distribution and postmortem considerations.
| Specimen or test | What was directly measured | What the result does not establish by itself |
|---|---|---|
| Femoral postmortem blood | Ethanol concentration by headspace GC | Premortem drinking, precise concentration while alive, fermentation quantity, impairment or drowning cause |
| Postmortem urine | Ethanol concentration in bladder urine when sampled | Absence of very recent drinking when negative; exclusive premortem origin when positive |
| Vitreous humor | Ethanol concentration in the sampled ocular fluid when available | A guaranteed correction to blood or definitive exclusion of all postmortem synthesis |
| Casewise cross-matrix pattern | Agreement or disagreement among measured concentrations | A direct assay of the microbe, fermentation mechanism, original dose or personal functional performance |
The investigators found no cases with positive blood but negative vitreous among cases with vitreous available. They identified four positive-blood/negative-urine cases, but no vitreous measurement for those four. In the other 619 blood-positive cases, urine or vitreous was also ethanol-positive, a pattern suggesting premortem ingestion without proving it in every case: postmortem production is not impossible in urine or vitreous under particular circumstances.
The authors did not calculate urine-to-blood ethanol ratios in this analysis. An unreported ratio should not be reconstructed or presented as if it were measured and validated in the study. For complementary matrix-specific principles, see vitreous-humor toxicology and blood-alcohol chromatography and analytical limits.
Does long submersion prove ethanol was created after death?
No. Duration of submersion, temperature and decomposition conditions can alter the opportunity for postmortem microbial processes, but elapsed time is not a laboratory identification of fermentation. Among 52 victims submerged for seven days or longer, the investigators reported that 50 did not show their defined cross-matrix evidence of postmortem ethanol production. Conversely, one of the four suspicious cases involved less than a day of submersion. The four cases spanned submersion periods from less than one day to 50 days.
The study found some statistical differences in submersion-time distributions across blood-ethanol categories. Those comparisons cannot tell whether microbes made ethanol in a specific specimen. A correlation between recovery interval and a laboratory category is not a calibrated case-specific microbial-production model.
Why can blanket exclusion rules bias alcohol-and-drowning research?
Removing every case with an extended interval between death and testing can remove genuine alcohol exposure along with uncertain observations. In this Finnish dataset, the authors calculated that a rule excluding cases beyond a 24-hour interval would remove at least 20.5% of alcohol-positive drownings. The corresponding exclusions were approximately 12.4% using a 48-hour cutoff and 11% using 72 hours. These percentages describe this specific cohort and its recorded times; they are not safety thresholds for accepting or rejecting an individual ethanol measurement.
Likewise, dropping every blood result below 50 mg/dL would discard all 28 low-positive cases, despite only four being flagged by the investigators’ urine-discordance screen. At least some low positives could represent actual recent ingestion. The reverse error is also possible: simply accepting every low-positive blood result as premortem exposure may overlook artifact.
The better epidemiological approach is transparent matrix-based evaluation and sensitivity analysis instead of presenting an arbitrary recovery-time or low-BAC cutoff as scientific proof of ethanol’s origin. The approach used for population classification should not be mistaken for a rule deciding an individual civil or criminal case.
Does a positive postmortem BAC establish impairment or drowning causation?
No. Laboratory concentration, biological exposure, functional impairment and event causation are different questions. The study contains no standardized task testing shortly before drowning, no individualized functional baseline, and no controlled experimental allocation of drowning risk. A measured postmortem BAC—even one whose premortem origin is supported by urine or vitreous results—does not demonstrate precisely what a person could see, judge, swim, coordinate or decide at the relevant time.
Nor does an association in population-level drowning data decide whether alcohol was a necessary or sufficient cause of a particular death. Environment, flotation, water conditions, underlying illness, activity, trauma, and witness evidence can matter. This study was designed to evaluate possible misclassification of alcohol exposure in drowning surveillance, not to produce an individual impairment threshold or probability of drowning causation.
What additional evidence is needed in a disputed case?
A forensic interpretation should begin with the actual collection record: specimen identity and site, time of recovery and autopsy, preservative, refrigeration or storage history, chain of custody, analytical method, replicate or confirmatory results, relevant detection limits, and decomposition indicators. It should establish exactly which of blood, urine and vitreous was collected and analyzed, instead of presuming all three exist.
Where the origin of ethanol is contested, case-specific microbiological testing, markers of putrefaction, ethanol metabolites and controlled fermentation work may sometimes inform the differential explanation. The Pajunen study recommends considering such additional evidence but did not validate those tests as a universal algorithm or retrospectively measure them in all 967 cases. Clinical history, documented recent drinking, witness accounts and the circumstances of the drowning are separate sources of evidence requiring independent scrutiny.
For background on appropriate postmortem sampling and interpretation, see postmortem toxicology principles. An expert should identify whether each conclusion rests on an assay, an observed circumstance, a population association or a conditional inference. That prevents a measured ethanol peak from being mistaken for proof of its source or its role in the event.
Frequently asked questions
Can microbes produce alcohol after a drowning?
Yes, postmortem formation is biologically possible. This retrospective study identified only four cases with a suspicious blood-positive/urine-negative pattern, but it did not demonstrate microbial synthesis in those individual cases.
What does blood ethanol of 25–48 mg/dL mean after death?
It is the measured concentration in the sampled postmortem blood, equal to 0.025–0.048 g/dL. The result alone does not distinguish very recent premortem drinking from possible postmortem formation or quantify the person’s functional state.
Does a negative urine alcohol test prove blood alcohol is from fermentation?
No. Urine may remain negative during sufficiently recent drinking, even though blood is positive. Collection and analytical limitations must also be considered. All four urine-discordant cases lacked a corresponding vitreous result.
Does positive vitreous alcohol always prove alcohol was consumed before death?
No. Vitreous is often helpful as a distinct matrix, but the study cautions that certain postmortem circumstances can also affect urine or vitreous. An individual conclusion depends on more than one finding.
Was 0.4% the proportion of blood-positive drownings caused by postmortem fermentation?
No. Four cases represented about 0.4% of all 967 selected drownings, about 0.64% of 623 blood-positive cases, or 14.3% of the 28 low-positive cases. None were experimentally confirmed as fermentation cases.
Does recovery more than 24 hours after drowning invalidate a BAC result?
No. In this cohort, a blanket 24-hour exclusion would have removed at least 20.5% of alcohol-positive cases. Recovery time is a circumstance to evaluate, not a validated binary rule for ethanol origin.
Can this research establish that alcohol caused a particular drowning?
No. It evaluates prevalence classification and possible postmortem ethanol production. Case-specific impairment and causal attribution require independent physiological, toxicological, factual and circumstantial evidence beyond the retrospective study.
Primary scientific source and practical conclusion
Pajunen T, Vuori E, Lunetta P. Epidemiology of alcohol-related unintentional drowning: is post-mortem ethanol production a real challenge? Injury Epidemiology. 2018;5:39. https://doi.org/10.1186/s40621-018-0169-4.
Forensic takeaway: Analyze the reported blood result as a concentration measurement; consider urine and vitreous as separate matrices; treat discordance as a reason to investigate rather than a definitive fermentation diagnosis; and never convert a population-level count into individual proof of drinking, impairment or cause of death. For drowning surveillance, overly broad low-concentration and time-since-death exclusions can themselves produce serious selection bias.