Urine as a Biological Specimen for Forensic Analysis of Alcohol and Variability in the Urine-to-Blood Relationship

Alan W. Jones

This article concerns the use of urine as a biological specimen for determination of alcohol in clinical and forensic toxicology and discusses factors that might influence variability in the urine/blood concentration ratio of alcohol. A large number of human drinking experiments were conducted to determine the time course of urine-alcohol concentrations (UAC) in relation to blood-alcohol concentrations (BAC). The UAC and BAC curves were shifted in time and the BAC curve always began to decrease before the UAC started to decline. During the early absorption phase the UAC/BAC ratio was less than unity, whereas in the late absorption/distribution period the ratio was between 1.0-1.2. On reaching the post-absorptive phase, the UAC always exceeded BAC and UAC/BAC ratios averaged 1.3-1.4, increasing appreciably as BAC decreased towards zero. Alcohol-induced diuresis was most pronounced during the rising portion of the BAC curve and near to the peak value. After about 2 hours post-drinking, the production rate of urine diminished to the pre-drinking rate of about 0.5-1 mL/min. Drinking water during the post-absorptive phase of the alcohol curve produced dilute urine, as reflected in lower creatinine content and osmolality, although the concentration of ethanol remained unchanged. After subjects drank a moderate dose of ethanol (0.54-0.85 g/kg) about 2% of the dose was recoverable in the urine after 7 hours. Ethyl glucuronide, a minor metabolite of ethanol, was measured in urine samples from drunk drivers. The UAC/BAC ratio of ethanol in drunk drivers did not depend on the creatinine content of the urine and therefore the relative dilution of the specimens. When alcohol-free urine was spiked with glucose and infected with the yeast species Candida albicans, ethanol was produced by fermentation after approximately 24 hours storage at room temperature. This post-sampling synthesis of ethanol was prevented by sodium fluoride (1% weight by volume) in the urine tubes or by keeping the specimens in the cold (4 degrees C). The UAC and BAC were highly correlated (r > 0.95) in drunk drivers and in autopsy cases, although the residual standard deviations were appreciable. This speaks against attempting to estimate BAC indirectly from UAC in any individual case. The UAC/BAC ratio and the change in UAC between two successive voids can help to resolve whether a large amount of alcohol had recently been consumed. This information is useful to support or challenge allegations of drinking alcohol after driving, which has become known as the hip-flask defence.