Vitreous Ethanol and BAC Conversion: Correlation, Prediction Intervals, and Uncertainty

Prepared by Okorie Okorocha, J.D., M.S., M.S.

Quick answer: Vitreous humor is a useful corroborating or alternative specimen for postmortem ethanol analysis, especially when blood is unavailable, decomposed, contaminated, or collected from a questionable site. But a strong blood-vitreous correlation across many cases does not authorize an exact conversion for one individual.

Why vitreous humor is analytically useful

The vitreous body is a water-rich gel within the eye. Its anatomical separation from the gastrointestinal tract and central organs reduces some risks of direct contamination and local diffusion. It can also remain available when trauma, exsanguination, decomposition, or embalming makes blood unavailable or difficult to interpret.

Those advantages are relative, not absolute. Vitreous concentration still depends on the phase of alcohol distribution, postmortem interval, collection technique, eye sampled, pooling, storage, analytical method, and case circumstances.

Correlation and prediction are different statistical claims

Two measurements can be highly correlated across a study population while still differing enough in individual cases to make precise conversion unreliable. Correlation describes how values move together. Prediction asks how closely an unknown individual blood value can be estimated from a measured vitreous value.

The 345-case Pounder and Kuroda study

Pounder and Kuroda examined paired vitreous and blood ethanol results from 345 medicolegal autopsies. Their regression described the group relationship, but the individual prediction interval was too wide for practical use. They also reanalyzed prior datasets and warned against conversion factors that omit prediction uncertainty.

The 706-case Jones and Holmgren study

Jones and Holmgren reviewed 706 forensic necropsies. Their principal ratio analysis included 672 paired results after exclusions. The arithmetic mean vitreous-to-femoral-blood ratio was 1.19, but the reported 95% range was 0.63 to 1.75. Log-transformed analysis produced a geometric mean ratio of 1.16 with 95% limits of agreement from 0.71 to 1.89.

Those data show why dividing every vitreous result by an average ratio can create false precision. The mean is not the individual case.

Smaller studies support usefulness without eliminating uncertainty

De Martinis and colleagues studied 21 ethanol-positive cases and reported correlations of 0.97 between vitreous and femoral blood or urine, 0.96 for heart blood, and 0.94 for subclavian blood. A later 62-case study reported a Spearman correlation of 0.96 between vitreous and femoral blood. These results support the value of vitreous as a companion or alternative matrix under the studied conditions.

Study differences matter: sample size, inclusion criteria, decomposition, climate, postmortem interval, preservation, sampling sites, and instrumentation vary. A regression from one dataset should not be imported into a new case without testing applicability and reporting uncertainty.

Absorption, elimination, and postmortem artifacts

Vitreous and blood do not equilibrate instantaneously. In an early absorption phase, blood can exceed vitreous. After distribution or during elimination, vitreous may exceed blood. The relationship is also vulnerable to non-pharmacokinetic explanations.

An experimental five-cadaver study placed ethanol- and drug-containing material in the trachea to simulate aspirated vomitus. After 48 hours, the highest ethanol concentrations occurred in pulmonary vessels, while femoral values remained low. This model does not supply a correction for ordinary deaths; it demonstrates that local diffusion can create site-specific central blood findings.

Other issues include central-blood contamination by gastric contents, putrefactive formation or loss, trauma, body position, resuscitation, transfusion, delayed collection, and specimen storage.

What EtG and EtS add

Ethyl glucuronide (EtG) and ethyl sulfate (EtS) are direct ethanol metabolites. Their presence may support antemortem alcohol consumption when ethanol source is disputed.

Thierauf and colleagues measured ethanol, EtG, and EtS in blood, urine, and vitreous humor from 26 cases. EtG was found in 24 vitreous samples and EtS in 23. Rainio and colleagues studied 58 autopsy cases, using a vitreous EtG immunoassay and LC-MS/MS analysis of EtG and EtS as the reference. In that sample, vitreous EtG was more sensitive than blood ethanol for detecting a documented history of excessive alcohol use, but positive and negative exceptions remained.

EtG and EtS do not answer every question. Their absence can reflect timing, elimination, degradation, method sensitivity, or case condition. Their presence supports consumption but does not independently quantify drinking, impairment, or cause of death.

Collection and laboratory records to review

• which eye was sampled, whether both eyes were pooled, sample volume, and visible contamination;
• needle placement, collection order, tube, preservative, headspace, seal, and labeling;
• death, refrigeration, autopsy, collection, shipment, receipt, storage, and analysis times;
• storage temperature, freeze-thaw cycles, and repeat-analysis history;
• headspace GC method, columns, internal standard, calibration, blanks, controls, duplicate criteria, and separation from other volatiles;
• EtG/EtS method, cutoffs, confirmation, ion criteria, matrix validation, and stability evidence;
• all blood sites, urine, vitreous, gastric contents, decomposition markers, and scene evidence.

How to evaluate a proposed conversion

1. Identify the original study population and sample size.
2. Confirm that the same specimen definitions, collection conditions, and method apply.
3. Require the residual error and individual prediction interval—not merely the correlation coefficient or regression line.
4. Test how the conclusion changes across the plausible interval.
5. State that the output is an estimate, not a measured BAC.
6. Do not use a converted value to imply more precision than the source data allow.

For the broader matrix discussion, see Vitreous Humor Toxicology. A litigation checklist is available at Vitreous Alcohol Evidence in Court.

Frequently asked questions

Can vitreous ethanol be divided by 1.19 to obtain BAC?

Not as an exact individual value. The published ratio had wide case-to-case variation.

Does a high correlation prove interchangeability?

No. Correlation does not establish agreement or narrow individual prediction error.

Is vitreous immune to postmortem change?

No. It is comparatively protected, but collection, timing, storage, diffusion, and analytical factors still matter.

Can EtG or EtS prove intoxication?

No. They can support antemortem consumption; intoxication and causation require additional evidence.

Selected sources

• Pounder DJ, Kuroda N. Vitreous alcohol is of limited value in predicting blood alcohol. Forensic Science International. 1994;65:73-80.
• Jones AW, Holmgren P. Uncertainty in estimating blood ethanol concentrations by analysis of vitreous humour. Journal of Clinical Pathology. 2001;54:699-702.
• De Martinis BS, et al. Alcohol distribution in different postmortem body fluids. Human & Experimental Toxicology. 2006;25:93-97.
• Thierauf A, et al. Ethyl sulphate and ethyl glucuronide in vitreous humor as postmortem evidence markers for ethanol consumption prior to death. Forensic Science International. 2011;210:63-68.
• Rainio J, et al. Immunoassay for ethyl glucuronide in vitreous humor. Forensic Science International. 2013;226:261-265.

Educational information only. It is not a conclusion about any individual case.