See also: John Bederka Toxicology Report

Toxicological Truths and Untruths

See also: Drunk Diving Defenses
See also:
Scientific Reliability
See also:
Toxicological Professional Societies
See also:
See also:
See also:
Kurt M. Dubowski
See also: Jury - Contains an expert witness directory http:/

  • Dr. Kurt M. Dubowski is the premier expert in blood and breath alcohol research today. (In actuality, Dr. Dubowski has published many articles on breath alcohol research but only three on blood and breath alcohol correlation's all of which were not peer reviewed.) Three articles on blood/breath correlation's from ALCOHOL TECHNICAL REPORTS, CLINICAL CHEMISTRY, and INTERNATIONAL MICROFORM JOURNAL OF LEGAL MEDICINE
  • Blood alcohol and breath alcohol measurements are interchangeable or equal. (They are two tests measuring alcohol in two different area of the body. These tests are based upon two completely different standards.)
  • Steepling or spiking phenomenon can be applied to blood alcohol testing in the absorption phase only.
  • Post drinking--a subject could have a rising blood alcohol content more than half an hour after cessation of drinking on an empty stomach.
  • A subject's blood alcohol could be .05 at 11:00 p.m. and when tested at 12:30 a.m. the test measured the BAC was .087 whole blood and .120 blood plasma. See also absorption rate. (This could happen only if the subject added more alcohol to his system.)
  • Widmark Model
  • Retrograde extrapolation or back calculation, cannot be performed on blood alcohol. (Widmark's research dealt only with blood alcohol on an empty stomach. Widmark never tested his theory on breath alcohol. Dr. Kurt Dubowski maintained for years that one cannot perform retrograde extrapolation on breath and blood alcohol. He has not made such a statement since the research conducted by Sajani and Dinn in 1985 when they tried to reproduce one of his two studies on blood and breath alcohol correlation's. Retrograde extrapolation is based on Dubowski's research. Interestingly, Dr. Herbert Moskowitz has written on this subjects for the U. S. Department of Transportation, National Highway Traffic Safety Administration. Dr. Moskowitz uses .017 per hour decline rate with the Total Body Water formula by Dr. Patricia Watson.

The Widmark Model was originally developed to explain blood alcohol equilibrium, absorption, and elimination.
Widmark, E. M. P.; 1932, Principles and Applications of Medicolegal Alcohol Determination, translated into English by R. C. Baselt, Department of Pathology, University of California, Davis, 1981, 163 p. There are three distinct phases after one imbibes alcohol--equilibrium, absorption, and elimination. With the first swallows of alcohol, all three processes begin. Widmark's work is based based only the BAC of a person whose blood has been tested for alcohol. Breath alcohol testing of the same person at the same time is approximately 10% -less than a blood alcohol concentration, Jones (1998, p. 337), Garriott (1996, 269-70) and Dubowski (1985, p.102).

"Widmark's method continues to dominate the way that forensic scientists and others deal with the pharmacokinetics of alcohol when requested to calculate the amount ingested from a single measurement of BAC, Jones (1998, p. 347). Jones (1984) used Widmark's to calculate BAC of 48 fasted men. 23 subjects peaked at 30 min., 14 at 60 min., 8 at 90 min., and 3 at 120 min. after start of drinking. 0.61 to 1.23 was the range for peak BACs. The elimination rate from blood was lower in those subjects with larger distribution volume. The results show that blood-ethanol parameters calculated according to Widmark's method have low inter subject variability when the dose of ethanol administered and the condition of the test subjects are carefully controlled.

"The work of Erik M. P. Widmark published in the 1930's is still widely accepted and cited in medicolegal practice when the pharmacokinetics of ethanol are discussed," Jones (1996, p. 99).

"This linear metabolic removal has been observed consistently, and it underlies the ability of long-established Widmark equations (A + p x r x Co ; and Co = Ct + Bt) to predict changes in blood alcohol levels from information on alcohol intake... In general, total body water (TBW) tends to average about 60 l for men and 40 l for women. Using anthropomorphic data (Watson,1981) circumvented using r and improved the accuracy in determining Co In this approach Co = .8 x A/TBW, and TBW is:

"TBW = 20.03 + .3626 x weight (kg) - .1183 x age (y r) for men.

"The literature of the past 50 years consistently described time-dependent changes in BAL following ingestion or infusion. Generally, oral ingestion gives an initial rise in BAL that represents absorption from the intestine, peaking usually within 1 h. Then there follows an approximately linear decline in BAL with time (zero order kinetics) equivalent to about 3 mmol per minute for people with about 50 l of body water which is based on data from Jones et al. and resembles results in other reports (Derr, 1993, Wallgren, 1970, and Wilkinson, 1977)," Lands (1998).

"If one had ever had a prescription drug, then Widmark's methods were applied for determining the dosage which is based on sex, weight, and age, etc.," (Lands, 1998.)

Alcohol equilibrium
"Experiments with intravenous injection of ethyl alcohol in humans have been carried out by Erich Gabbe. His determinations show that the value for the blood alcohol concentration which corresponds to approximate diffusion equilibrium in the entire organism is attained by 8 to 10 minutes in most cases and within 30 minutes in all experiments, (Widmark, page 16).

Amount of alcohol absorbed from the stomach and the small intestine
"Various authorities seem to agree, however, that fasting individuals absorb 20% to 25% of a dose from the stomach and 75% to 80% from the small intestine proximal to the pyloric sphincter. Negligible amounts are absorbed form the mouth and colon (Schwar, 1979)," Baselt, 1996, p. 66.


From ALCOHOL ALERT, 35, PH 371, January, 1997.

Blood alcohol concentration (BAC) after the rapid consumption of alcohol by eight adult fasting male subjects.* (Adapted from Wilkinson et al., JOURNAL OF THE PHARAMCOKINETICS AND BIOPHARMACEUTEICS 5 (3): 207-224, 1977.)

100 mg% is the legal level of intoxication in most States. 50 mg% is the level at which deterioration of driving skills begins. (JAMA 255: 522-427, 1986.)

*If the same number of drinks are consumed over a longer period of time, BAC's will be lower.

Time To Reach Peak Blood Alcohol
"Wallgren (1979:165) quotes a person absorbing from the gastro-intestinal tract in fasting conditions of about 1.5 to 2 hours in humans subjects and prolongation of the process to 6 hours or more in the presence of food. These figures are based on work published by Tuovenin in 1930 (!) performed with the same individual (!) after the ingestion of 60 ml of alcohol (5%, 20% and 40% concentrations) in fasting condition and immediately after eating 1,1 kg potatoes," (Cooper,Schwar, Smith, 1979, p. 91.)

"Cortot et al. showed that when alcohol was taken with food, 73% of the dose was absorbed through the stomach and only 24% was taken up through the duodenum. Indeed, some of the alcohol was retained in the stomach for as long as 6 h after drinking," (Jones, 1093p., 1994).

"According to AW Jones and his research done over the last 15 years, 77% of the subjects (n=152) reach peak BAC in less than 45 min after the end of drinking and 97% in less than 75 min with a bolus dose of .34-1.02 g/kg. After a dose of .80 /kg in the form of a 20% v/v orange juice cocktail, consumed within 30 min, peak BAC occurred within 60 min after drinking in 92% of the trials (n=65).

"Shajani and Dinn reported that the average time to peak BAC after the end of drinking was 35 min (range 17-68 min) when the alcohol was consumed as mixed drinks in moderate does and over longer drinking times. Gullberg tested 39 subjects who drank alcoholic beverages of their choice with and without food. He found that the mean time to peak BAC after drinking was 19 min (range 0-80 min) and 81% of the subjects peaked within 30 min. MaCallum and his collaborators reported beer drinking experiments which resulted in an average peak BAC of .132% w/v (range .04-.26% w/v). The mean time to peak BAC after the end of drinking was 55 min and the range was 20-114 min.

"These results support the notion that after heavy social drinking over several hours the BAC profile should be postabsorptive 30 min after the end of drinking. In our studies on ethanol absorption, we found that the biggest increment in BAC occurred within the first 15 min after drinking. The rise is progressively less as the peak BAC is approached as expected for a first-order process," (Jones, 1990, p. 199).

"The pyloric sphincter retards the emptying time of the stomach when in spasm (Mueller, 1975: 1010) or where its opening is inhibited by high concentrations of alcohol. Various forms of physical obstruction as carcinomatous and fibrosis have a similar effect.

"Psychological states, e. g. emotional disturbances, influence the motility of the stomach and the pylorus sphincter. The effect is dependent upon the reaction which is elicited--sympathicomimetic or colinergic.

"Prior alcohol exposure, especially chronic alcohol intake, can affect alcohol metabolism. Chronic alcohol consumption generally increases the rate of ethanol elimination, in one study, an increase of 90% was reported (Bode, 1979). Similarly, Winek and Murphy have found that nondrinkers have slower ethanol elimination rates than those of social drinkers and alcoholics," (Chan, 1986, p.105).

"In contrast to most drugs, the blood ethanol clearance curve approximates linearity until very low concentrations are reached. The discovery of this linearity (or pseudo linearity) is traditionally ascribed to a 1919 report by Mellanby," (Sturtevant and Sturtevant, 1989, p. 27).

Montgomery's Rule of Thumb--"Metabolism in the liver and elimination by the kidneys remove .015 -.020 g/dl per hour by a zero-order process (Ritchie, 1985), which is constant within the range of BAC's normally encountered in forensic considerations. This hourly decrease in the BAC is roughly equivalent to the BAC produced in a 70 kg individual by absorption of one drink. Hence, a practical "Rule of Thumb" provides that individuals of average mass eliminate one drink per hour," (Montgomery ,1992).

"However, body mass does become relevant when part of the desired prediction is estimation of the number of drinks consumed. Three parameters are inter linked: (1) body mass, (2) number of drinks absorbed, and (3) blood alcohol level. If any two are known, the third can be calculated based on standard pharmacokinetic principles, where

BA + mass of alcohol consumed / volume of distribution x body mass.

Alternatively, the desired parameter may be obtained from Table 1 based on these calculations (Sunshine 1975). These combined principles result in the absorption and elimination of alcohol from a hypothetical individual as shown in Figure 1. This curve, which has been reported widely and is presented here schematically, has been determined repeatedly following bolus consumption and illustrates the characteristics of nonlinear absorption, peak concentration at 30-90 min following a bolus consumption, and linear elimination at a rate of 0.02 g/dl/h.

"These toxicological principles, as just presented, also may be used to describe blood alcohol kinetics in realistic situations. A hypothetical set of circumstances is presented for illustration: Ten drinks are consumed by a normal individual on relatively empty stomach during a 3-h period, after which no additional alcohol is ingested. The expected BAC's for three reasonable, but widely different, consumption patterns in the same individual are presented in Figure 2.

"Case A--five drinks are consumed during the first hour, three drinks during the second hour, and two drinks during the third hour.

"Case B--three drinks are consumed during the first hour, four drinks during the second hour, and three drinks during the third hour.

"Case C--two drinks are consumed during the first hour, three drinks during the second hour, and three drinks during the third hour.

"Considering a constant rate of removal of one drink per hour, which will occur simultaneously with absorption, all three situations will result in seven drinks remaining in the individual's body, but not necessarily absorbed into the blood, at the end of the 3-h drinking session. the differences amount the three cases in resultant BAC's at h 3 will reflect the number of those seven remaining drinks that have been absorbed. (But not eliminated.)

Figure 2 illustrated the differences between this "real life" situation and that situation resulting after bolus consumption (Fig. 1). The principal difference relevant to pharmacokinetics is that, in Figure 1, the peak absorption occurs 30-90 min after bolus administration. In figure 2 the time of peak absorption varies, reflecting the differences in drinking patterns during this absorptive state. However, in all three cases, a similar peak blood level is reached within 30-60 min after cessation of drinking (after h 3), and in all three cases the BAC's are virtually identical thereafter. These drinking patterns reflect more realistically what is encountered in the normal social environment. In fact, allowing an additional 30-60 min to complete absorption after the last drink may be overly conservative. The peak blood level most likely is reached simultaneously, or even before the last drink is consumed, in normal drinking situations (Gullberg, 1982; Jones, 1990)," (Montgomery, 1992).

Post-drinking and absorption
"The uncertainty in the amount of alcohol ingested was greater when estimated from the BAC at 1 h(our) post-drinking, compared with the BAC at later times. Widmark's equation rests on the assumption that all ingested alcohol is fully absorbed and equilibrated in body fluids and tissues at the time of sampling.

"If drinking extends over several hours this will reduce the likelihood of an overshoot peak after the last drink. Because the absorption of alcohol begins at the start of drinking, negligible amounts probably remain in the stomach at the end of drinking. Under these circumstances even the 1 h post-drinking BAC might give reasonably good estimates of the amount of alcohol ingested" (Gullberg and Jones, 1994, p128-9).

"Absorption of ethanol starts in the stomach where about 20 percent of the dose can become absorbed. The remainder is absorbed from the upper part of the small intestine. The peak BAC and the time of reaching the peak after drinking are important aspects of the absorption kinetics. The absorption of ethanol occurs progressively during a drinking binge or spree, and studies have shown that the BAC fifteen minutes after the last drink has reached about 80 percent of the final peak BAC," (Jones, 1995, p.805.)

Gullberg evaluated the time course change in blood alcohol concentration (BAC) specifically following the last drink and on a empty stomach. "39 volunteers submitted to consuming alcoholic beverages under eight different drinking conditions, which involved varying the consumption time, volume consumed, and food content of the stomach. The results indicated that among the eight different drinking conditions, there was not a statistically significant difference in BAC change or time to reach the BAC plateau. All of the drinking conditions produced near similar results. All of the drinking conditions produced near similar results. The most significant result was that the rise in BAC following the last drink was .02 g/100 ml or less for 81.3 percent of the volunteers and that 81.3 percent of the volunteers also achieved their maximum BAC within 30 minutes or less. The maximum individual rise in BAC was .04 g/100 ml, while the maximum individual time to reach the BAC plateau was 80 minutes. The results would suggest that regardless of variations in consumption time, volume consumed, or food consumption, the BAC plateau is reached rapidly following the last drink and does not rise significantly as suggested, the statistical t-test for paired data). The molecular nature of ethanol and the structure of the gastrointestinal tract would also suggest this observation to be the case.

"The results suggest the low probability that an individual submitting to a blood alcohol test following an incident will have a BAC that is higher than at the time of the incident. This is particularly true if a period of time has already elapsed between the time of the last drink and the time of the incident. The question as to whether the defendant is increasing in BAC at the time of the incident can be further clarified with through investigation by the law enforcement officer. Asking key questions along with careful documentation will be invaluable for court presentation," Gullberg, 1982).

"For some unexplained, and physiologically improbable reason, the alcohol ingested during the evening remained unabsorbed in the stomach until the person decided to leave for home or dive to the next bar. Shortly after driving the person is either involved in an accident or pulled over by the police because of a moving traffic offense, and in this connection is arrested for DUI. The defendant then claims that between the time of being apprehended and the time of taking the blood or breath-alcohol test, the alcohol in the stomach has become absorbed into the blood bringing the person over the legal limit.

"Obviously, this scenario is unreasonable because alcohol, unlike many other drugs, starts to become absorbed from the stomach immediately following ingestion. Gastric emptying accelerates this process and leads to a rapid onset of the effects of alcohol on the brain. Indeed, people indulge in drinking primarily to experience alcohol's enjoyable pharmacological effects such as euphoria, relaxation and diminished social inhibitions. In order for this to happen, the alcohol must become absorbed into the blood and transported to the brain. The intoxicating effects of alcohol are more pronounced during the rising limb of the BAC profile, and people would surely be surprised if they had been consuming drinks for several hours without experiencing any effect! Unfortunately, only a handful of studies have looked at the pharmacokinetics of alcohol under real world drinking conditions to establish, for a large number of subjects, the degree of rise in BAC and the time needed to reach the peak after the last drink, (Jones, 1998, p1138.)

Gullberg reported a study in which 39 subjects drank various quantities of alcohol under real world drinking conditions. The mean time required to reach the peak BAC after the end of drinking was 19 min (span 0 to 80 min) and 81% of subjects reached a peak within 30 min. A study reported by Shajani and Dinn also give a clue to the time needed to reach peak BAC under social drinking conditions. In 8 men and 8 women who consumed known amounts of alcohol according to choice, the maximum BAC was reached in 35 min (span 17 to 68 min) after end of intake. Taken together these studies and a few other suggest the low probability that the result of a blood or breath-alcohol test made some time after driving will be higher than at the time of driving which is often 1 to 2 hrs earlier. Zink and Reinhardt made an important contribution when they allowed heavy drinkers to consume very large amounts of alcohol over periods of 6 to 10 hr, resulting in peak BAC's in the range of 0.10 to 0.38 g/dL, and taking samples of venous blood for analysis of alcohol at 15 min intervals during and after the drinking spree. In this way accurate information was obtained about the shape of the concentration time profile and the time of reaching the peak as well as rise in BAC after the last drink. Importantly, they found that half the individuals had reached their peak BAC even before the last drink was taken (I.e., the rate of elimination exceeds the rate of consumption). The longest time necessary to reach a peak was 50 min after the end of drinking (mean + SD, 7.7 + 22.9 min), and when a rise BAC occurred between the end of drinking and the peak BAC it was invariably less than 0.02 g/dL. This study has important ramifications because many DUI suspects have blood-alcohol concentration in this high range when they are apprehended," Jones, (1998, p. 1011).

Steepling--a phenomenon of breath alcohol testing
"The declining limb of the blood alcohol curve is not a straight line. This is demonstrated in the post-peak phase of Alha's curves (1951:38-41), and in the differences in the B 60 and 60 values. With the use of more sensitive analytical methods, Shumate et al (1967: 83-100), with the Breathalyzer (Figure 9) and Wehner (1972: 81-93) with the direct recording auto-analyser, demonstrated these irregularities more clearly (Figure 10). Similar results were obtained by Hansen & Disse (1962: 117-26) with analyses of specimens taken at six-minute intervals.

"With regard to the fluctuations found in their experimental work, Shumate et al conclude that it is difficult to believe that the observed fluctuations can be accounted for by variation in the rates of alcohol eliminated for a given individual within a given test. This would require the body to have some control mechanism which would compensate almost perfectly for periods of high elimination by subsequent period of low elimination and vice versa. A much more credible alternative hypothesis is that the variances of errors in observation account for most of the observed fluctuations. The instrument used had graduations of 0,01%. Since any measurement less than 0.01% requires visual interpolation by the technician, the variance of the reading error could easily be substantially larger than the variance in the rate of elimination for a 5-minute interval. "A striking feature of the curves, especially those of Shumateet al. (1967: 83-100) is that in spite of the irregularities or oscillations, the four curves run practically parallel with each other. A similar feature is seen in Alha's curves," (Cooper, Schwar, Smith, 1979, p. 232). Note: this is breath alcohol.

"An important challenge to the forensic issue of breath alcohol analysis is the so called "steepling effect", large positive and negative excursions in short time intervals over the course of an individual's breath alcohol concentration (BrAC) time curve, (Mason and Dubowski, 1976). Concentration time curves appear noisy, with peaks and valleys (hence, the term "steeping") over time. The steepling phenomenon must be attributable to combined analytical and biological components inherent in breath alcohol sampling and measurement. Moreover, when discussing the issue of steepling, one should always provide a numerical estimate of its magnitude, such as RSS or Sy/x. Interpreting the phenomena as originating from some other biological cause (e.g. pyloric spasms, etc.) is cautioned against unless total analytical variability is accounted for," (Gullberg, 1994, p. 321).

"The 'steepling' phenomenon is due largely to sampling variability and is capable of being reduced to acceptable levels with adequate attention given to sampling criteria and procedural technique. Any report of 'steepling' in breath or blood alcohol kinetics should have associated with it some numerical assessment of its magnitude and source," (Gullberg, 1998).

"A zig-zag concentration-time profile of ethanol was demonstrated when end-expired air was used as the biological specimen for analysis of ethanol. There were inter- and intra-subject variations in the amplitude of spiking, being most pronounced during the absorption phase...Short term fluctuations in BAC are more likely to occur during the absorption phase of ethanol kinetics. This probably reflects periodic evacuation of the stomach contents into the intestine and variations in peripheral blood flow. Therefore, the zig-zag effect does not reflect short term fluctuations in the rate of ethanol metabolism. Instead, we consider that the sporadic fluctuations reported by some workers are most likely caused by pre-analytical sources of variation such as those associated with sampling blood or breath for analysis," (Jones, 1990).

"Figure 4-6 shows blood-ethanol concentration-time data from an experiment described by Teige et al, (1974). Samples of venous blood were taken every 1-2 minute through an indwelling catheter and ethanol was determined by gas chromatography. When plotted as a function of time an irregular zig-zag pattern is clearly seen. But this zig-zag trend can be accounted for by a combination of pre-analytical and analytical variations inherent in the methods of sampling blood and determination of alcohol. The insert shows the blood-ethanol time course for samples taken every 20-30 min. The zig-zag effect is now obliterated which underscores the importance of sampling protocol when studying this curious steepling phenomenon. In conclusion, it seems that the zig-zag patterns in blood-and breath-alcohol profiles are mainly the consequences of pre-analytical and analytical variations in the methods used for blood-alcohol analysis or breath sampling parameters when breath-instruments are used and breath-to-breath physiological variation in alcohol concentration. However, during the absorption of ethanol from the stomach an irregular rising phase has often been observed owing to the sudden and unpredictable opening and closing of the pyloric sphincter, a muscle that controls emptying of the stomach contents and the duodenum. This might produce a series of short bursts in absorption of alcohol into the portal blood flowing to the liver resulting in a sudden rise in the peripheral blood-alcohol concentrations," (Jones,1996, p.109-112).

"The technically more difficult problem of continuous monitoring of ethanol in expired alveolar air, further to confirm "steepling", appears not to have been undertaken to date," (page 29). " Data are not presently available to permit an opinion as to whether the steepling effect presents a forensic problem for breath analysis, except in the comparison of results of analysis for ethanol in nearly simultaneous blood and breath specimens," (Mason, 1976, p.32).

Absorption Rates
"The rate of absorption also varies from person to person. Generally, most alcohol is absorbed in 20-30 minutes in a fasting person after a single dose. Nearly 90 percent is absorbed within one hour. When drinks are consumed successively over time, the blood alcohol levels rise with each drink, reaching a maximum in about 20 minutes, then the levels decline from this maximum.

"It has been established that ethanol is eliminated from the body by principally by metabolism and a minor extent by excretion. The rate of metabolism varies from person to person; the normal range is 7.5 to 10 gm of ethanol per hour. Based on an average of 50 liters of body water, the above mentioned relate of metabolism would cause a blood alcohol concentration to decrease by .015 to .02 percent per hour. The rate of elimination is essentially linear, so that for any given person's alcohol elimination per hour will be constant.

"A fasting subject drinking a 20 percent solution of a carbonated alcoholic beverage will have a higher and earlier BAC maximum. When several drinks are consumed in a very short period, peak BAC's may not appear until about 45 minutes or longer, after the last drink. And finally, when alcohol is consumed successively over a period of time, the blood alcohol curve, will rise with each drink and then decline as the alcohol is eliminated.

"The time elapsed since the first drink must also be considered, since alcohol is continually being eliminated from the body. Although one's individual elimination rate is not known, the average value is .018% per hour. This figure times the number of hours since the first drink is the amount of alcohol eliminated. Furthermore, different individuals at the same BAC may react differently. Persons accustomed to alcohol use will learn to compensate for some of the effects of alcohol and in turn perform certain acts better than those unaccustomed to drinking. For example, they learn to minimize swaying by standing with their feet apart and they gradually learn to walk a straight line when mildly intoxicated. It is however, the opinion of most investigators that tolerance is limited and occurs at blood alcohol levels in the lower levels of intoxication as opposed to higher levels," (Winek, 1985, p. 34-61.)

"At all but very low does of alcohol, the live alcohol dehydrogenase is saturated, and after the peak BAC has been reached, the decline of BAC follows a fairly constant rate of about .015% h. This rate is independent of dose.

"al-Lanqawi measured 24 fasting male volunteers for up to 9 hrs after a single oral dose of 710 mg kg. In each individual, three elimination rates were used to back-extrapolate plasma ethanol concentrations over 3 and 5 h periods from observed values at 4 h and 6 h post-dosing assuming zero-order kinetics. The extrapolated values were then compared with the observed concentrations. When a ko (elimination) value of .015 mg l h (a value often cited as a population mean) was used for back extrapolation this resulted in significant underestimation of actual values and therefore would be unlikely to result in valid prosecutions," (al-Lanqawi,1992).

"On average, the rate of metabolism of alcohol for adults is 15 mg ethanol/100 ml blood/hour," (Fisher, 1987, p. 301).

Rising Blood Alcohol
"Apprehended drunk drivers sometimes plead a rising blood-ethanol concentration (BEC) as their defense. This defense rests on the assumption that the BEC was below the statutory limit at the time of the offense but above the limit when specimens of blood or breath were obtained for quantitative determination of ethanol. The status of ethanol absorption in drunk drivers is an important consideration when attempts are made to estimate the BEC at the time of the offense for the BEC determined at the time of sampling, which is often a few hours later. This technique is called backtracking BEC, or retrograde extrapolation. Expert testimony on both these issues requires careful consideration of the absorption kinetics of ethanol and the factors influencing this process.

"Jones (1991) et al. studied 152 men aged 20 to 60 and their weights were 60 to 109 kg. They arrived at the laboratory at about 7:30 am without eating breakfast. They started to drink ethanol at about 9:00 am and the duration of intake was 15, 20, 25, or 30 min, depending on the dose. The tests were conducted on the same subjects on two occasions 14 days apart and the results showed good agreement between the two tests with the same dose of alcohol.

"The times needed to reach the peak after drinking were not normally distributed; the mode was 30 min after the start of drinking, corresponding to between 5 and 15 minutes after the end of drinking. A total of 77% of the subjects reached their peak BEC between 0 and 45 min after the end of drinking and 92% between 0 and 75 min. after drinking. Five subjects (3.2%) reached peak BEC between 95 and 105 min after drinking. In the post peak phase, however, the BEC in both subjects agreed within the 95% confidence interval calculated according to Widmark's equation.

"We have demonstrated that the absorption profile of ethanol on a fasting stomach is reproducible in the same individual at least over a period of several weeks or months," (Jones, 1991, p. 376-385).

"The experiments by Shajani and Dinn indicate a mean of 35 min as the time to reach peak after the end of drinking (range 17 to 68 min) when ethanol was consumed as mixed drinks for longer periods. Gullberg made tests with 39 men who were allowed to drink various doses of ethanol with and without food. He found a man time before the peak BEC was reached of 19 min (range 0 to 80 min). Zink and Reinhardt allowed subjects to drink large quantities of ethanol over a period of 4 to 10 h. Interestingly, for 8 of 14 subjects the peak occurred either before or at the same time as the last drink was finished (mean, 14 min; range, 0-56 min); the last drink after this high initial ethanol consumption failed to increase the venous BEC. In the remaining six subjects, the peak BEC occurred, on average 22 min after the drinking ended (range 10 to 50 min). These results support the notion that after heavy social drinking, the BEC time profile has probably passed the peak and is decreasing by the time the specimens of blood are obtained for legal purposes," (Jones, 1991, p.376-385).

"Calculations of blood alcohol concentrations (BAC) by forensic alcohol specialists are routinely requested in court proceedings, given an amount of alcoholic beverage consumed or as a result of retrograde extrapolation from a preexisting BAC. These calculations when conducted are always qualified by the authors, with the assumptions that: (i) there was no consumption of alcohol for at least 30 min prior to the time of interest; (ii) and that the rate of elimination used for those calculations is between 10-20 mg/100 ml/h." Shajani and Dinn tested 16 subjects, nine who had alcohol with a meal and seven who consumed alcohol after a three to four hour fast. Both blood and breath samples were taken. "The mean time to reach a peak BAC after the last drink for Group I subjects (n=5) was found to be 36 min, with a range of 12-74 min. Similarly, the mean time to reach a peak BAC after the last drink for Group II subjects (n=9) was found to be 35 min with a range of 15-74 min. Although the consumption of food is reported to delay the absorption process of alcohol, no delay effect in the time to peak BAC was noted in this study.

"Schlayer and Lechner, in an analysis of 1432 blood alcohol curves extracted from the literature and other sources, came to the same conclusion that food ingestion did not cause a significant delay in the appearance of the blood alcohol peak," (Jakus, Shajani, and Image 1992).

Santamaria Chart and Ethanol Ingestion Studies

This chart appears on page 8:27 of Lawrence Taylor's Drunk driving defense, 4th edition, Little, Brown and Company, c1996. This chart is from Santamaria, J. N., Ethanol ingestion studies, Department of Community Medicine, St. Vincent's Hospital, Fitzroy, Australia Department of Transport, Office of Road Safety, 1979.

The Australian Medical Student's blood alcohol peaked at 14 minutes after drinking ceased, he had been drinking for two hours (the two hours plus). Absorption is still going on, it is not the dominant process.

Note: Spiking occurs in the absorption period, when the blood alcohol is known to fluctuate. The breath alcohol line (represented by the broken line) shows spiking after the BAC peak has been reached, which was 2 hs. and 14 min from the start of the experiment, blood alcohol does not appear to spike. The breath alcohol line spikes during the absorption and elimination phases, (Santamaria, 1979, Experiment 2, Subject B).

The subject was a male medical student, weight 69.6 kg, "Experiment 2: Alcohol was administered as in Experiment 1: (five 200 ml glasses of beer (equivalent to 38 grams of ethanol) were administered at 12 minute intervals to four fasting male subjects. The final drink was consumed within the total time of 60 minutes. In this and subsequent experiments in which fasting subjects were used, a minimum period of 8 hours without food or drink was demanded.) Alcohol was administered as in Experiment 1 for the first hour. In the second hour, two further 200 ml glasses of beer were consumed at 24 minute intervals. Thus a total of 53.2 grams of ethanol was consumed over a two hour period."

"Blood Samples: split samples of blood were obtained for blood analyses. An indwelling needle was inserted into a vein in the cubical fossa, and the lumen was kept patent by a solution of heparin-saline. Samples were collected into fluoride oxalate tubes. Additional samples were collected into lithium heparin tubes for separation of plasma to be stored in a deep freeze. Early samples were drawn at five minute intervals. Gas chromatographic tests on blood samples were conducted by the Forensic Science Laboratory of the Victoria Police.

"Breath Samples: two Breathalyzers (Model 900) were provided and operated by the Victoria Police. These were standardized by the breath analysis section of the Victoria Police and operated by trained field officers of the squad. In most cases, breath analysis was carried out every 10 to 20 minutes, commencing about 20 minutes after drinking had ceased.

"Carlton Draught Beer containing 3.8% w/v ethanol (7.6 g/200 ml) was used." Santamaria, JN, "Ethanol Ingestion Studies", Department of Community Medicine, St. Vincent's Hospital, Fitzroy, Australia Department of Transport, Office of Road Safety, 1979.

The study shows the fluctuations in the absorption phase of the alcohol elimination curve. That area (absorption) is well known for its fluctuations. Test 2 Subject B's BAC peaked approximately 14 minutes after the last drink and all four of the subjects in the test had erratic BrAL's previous to the peak and after the peak when the breath alcohol was measured. Subject G, a female, in Experiment 4 had the least fluctuation in the BrAL's, (Santamaria, 1979).

An approximate estimation of the length of the absorptive period in the experiments where determinations were performed in this part of the alcohol curve gives these results: when 30-50 g of alcohol is taken in a volume of 100-1000 cubic centimeters on and empty stomach, the absorptive period lasts 50-80 minutes in most cases. No period longer than 110 minutes has been observed", (Widmark, 1981, p. 64). (It has never appeared in scientific literature in this country or Australia. There are many articles in ETOH by Dr. Santamaria, this is not one of them.)

Kurt M. Dubowski, PhD.
Retrograde Extrapolation and Blood and Breath Alcohol
"Using the BAC's from these samples, it was possible to fit a regression equation to BAC's estimated from breath and those determined by direct blood analysis. This equation was then used to adjust all of the breath results so that they were no longer underestimates of the actual BAC's. Consequently, all the BAC results presented are estimates of the actual BAC's that would be expected from analysis of samples and not under estimates of the BAC's that would normally be obtained by breath analyses employing a 2100:1 conversion factor." O'Neill, B.; Williams, A. F.; Dubowski, K. M., "Variability in blood alcohol concentrations: Implications for estimating individual results", JOURNAL OF STUDIES ON ALCOHOL (1983), 44 (2): 222-230.

"Whenever an individual blood alcohol clearance curve is determined, it is, in effect, a measurement of total body water, and all precautions applied in total body water measurements should be incorporated in the experimental protocol. Unless the purpose of the study is to determine the effect of food on the absorption and elimination of alcohol, all clearance curves should be measured after an overnight fast.

"In alcohol studies in recent years, this requirement has frequently been neglected, e. g., variations in elimination rate cannot be assessed if the results are confused by the effect of food on alcohol equilibration. Here the experimentally measured elimination rate and Co value may well differ from their real value", (Watson et al.,1989).

Dr. Dubowski has not written on the retrograde extrapolation aspect since 1985 and the Shajani and Dinn article. However, he continues to do research on breath alcohol and testing. Dubowski may also have had misgivings about how his work was interpreted by forensic experts such as Jensen. Perhaps that is why, he wrote the article on expert witnessing, (Dubowski, 1988).

Alcohol Technical Reports
Dubowski, Kurt M. "Human pharmacokinetics of ethanol I: peak blood concentrations and elimination in male and female subjects", Alcohol Technical Reports, 5 (4):55-63, 1976.
Journal is no longer published, ceased in 1984. THIS ARTICLE WAS NEVER PEER REVIEWED. There are no dates for submission and acceptance. THERE ARE MAJOR PROBLEMS WITH THE DATA IN THIS STUDY. Had it been peer reviewed, several of the errors would/should have been corrected. Dubowski proved that retrograde extrapolation works, however we do not know if it is breath or blood, but probably breath alcohol tests.

1) Most research begins with a review of the literature, or why the study is relevant.

2) The paper begins with "41 subjects", "27 male and 14 female" on page 55; on page 57 paragraph 1 there are "15 females" and Table 1 "N = 29"; page 58 Table 3 "Total # = 40", paragraph 1 states "41 subjects", paragraphs 2 states "40 subjects", paragraph 4 states "39 subjects"; page 59 Table 4 "Males (N = 25). Females (N = 15) total N = 40", Table 5 "Total # = 39", paragraph 1 states "39 subjects"; and page 60 Table 6 gives the "Male (N = 25)" and the "Female (N =1 4)" with a "Total (N = 39)", Table 7 at head of the column "This study N = 41" and paragraph 3 states "38 subjects"; and finally page 61 "N = 38". None of this is explained. Why are there so many differences in the numbers of persons tested? There should be some consistency in the data.

3) "Range for Elapsed Time is 15-138" but '14' is clearly at the top of the column so the range should be 14-138 (elimination rate of fed subjects) Table 3 on page 58.

4) Paragraph 4, page 58 states "From the Individual data for BAC versus time obtained by direct analysis of serial blood specimens, Co values were calculated by linear regression"... It states the regression is done with least squares, but that is under figure 2 and he doesn't tell us how the least squares were produced.

5) No indication was given of: length of fast before the test or if they were fasted at all, what constitutes a "light sandwich meal", if food was given to the subjects at another time, length of time to drink the alcohol, how many drinks was it divided into, what was the time of day of the testing, the target BAC or BrAC to be reached, where the study was conducted, i.e., a hospital, a laboratory, or a bar.

6) What was the interval between the tests--what does "frequent intervals" mean?

7) Were is the "nearly simultaneous" blood test information, breath information is only presented?

8) Page 58 in paragraph 3 author converts BAC to BrAC by using the "uniform density of 1.055 g/ml". Page 59 paragraph 1 the author tries to convert BAC to breath alcohol. This was based on a (conversion of BAL from mg/liter to mg/kilogram based on an assumed uniform blood density of 1.055 g/ml)this is applies to the conversion of BAC to BrAC and vice versa. Dubowski later writes that one can't convert BrAC to BAC. This is on page 102 or his article "Absorption, distribution and elimination of alcohol", JOURNAL OF STUDIES ON ALCOHOL, Supplement 1985, July, p. 98-108.

9) Page 60 "regression of y (=Peak BAC) upon x (=Elapsed Time) was r2 = 0.042." Looks like concentration x time x r. Paragraph 2 (from Table 4) "BAC decrease between 8.9 and 23.3 mg/dl1 hr1". But on Table 4 "10.6-22.1" is listed for the range. Where does he get r2 = 0.042, on page 61, r2 = .093.

10) When these considerations (uncertainty of presumed postabsorptive state, assumed linearity and regularity of BAC decrease with time) are coupled with the "steeple effect" in the BAC time curve and with the wide range of B60 (elimination rates) expectable, as demonstrated above, it becomes clear that speculative retrograde extrapolation in the BAC to any point prior to an experimentally determined value must be avoided in forensic practice. . .", page 61.

11) Dubowski states on page 61 "Our results for this value (Table 5) are somewhat higher than Widmark's own data in the smaller subject series, but coincide well with the values reported in several other studies shown in Table 8. Dubowski's Factor r is 0.73 for men and 0.66 for women. Widmark's r was 0.68 for men and 0.55 for women. Dubowski's were not fasted, did they eat later, too?

12) Page 62 paragraph 2, Dubowski states "they (the data) demonstrate the hazards attending predictions of expectable peak BAC's after social alcohol consumption of the time required for reduction of an initial BAC to a give lower values, as well as the limitations, in forensic practice, of speculative retrograde BAC extrapolations using an assumed universal rate of BAC decline. Therefore, the expert should give a range of BAC levels, not just one level." (Widmark maintained that a range should be given.)

13) Widmark's experiments were conducted on empty stomachs. Gullberg and Jones (1994) maintain you can't apply Widmarks to breath alcohol testing unless another source of variation, K (the blood breath conversion factor) is introduced.

14) Page 62 gives information about a "typical male with a BAC of 250 (.25) who takes 17 hours to reach the alcohol free status. If you divide .25 by 17 you get an alcohol elimination rate of .0147 (nearly the same as Table 59.) Dubowski proved that retrograde extrapolation works, however we do not know if it is breath or blood, but probably breath alcohol tests.

Dubowski, KM, "Human pharmacokinetics of ethanol," INTERNATIONAL MICROFORM JOURNAL OF LEGAL MEDICINE (1975), 10 (4): 1-16.

The previously mentioned journal article from ALCOHOL TECHNICAL REPORTS was based on this article. Such was noted at the bottom of the article as being a revised version. Certainly there were still many inconsistencies which will be noted:

Pages 1-2 verbatim.

Page 3: "(27 males, aged 21-50 years, and 14 females, aged 23-46 years)" is the only alteration.

Page 4 is verbatim.

Page 5: Table 2 all of the values have been changed except for the number of males (N=27), number of females (N=14) and Total (N-41). At bottom of page 5 "were obtained by direct analysis of blood specimens and were not derived from the results of breath-alcohol analysis" is omitted from the ATR document. Females S. D. + 21.6 was the same. All other values were higher except the C.V. values which were all lower, even though the number of subjects remain constant.

Page 6: Under Figure 2 "(The Least Squares Regression Equation for These Date is Y=110.5X + 3.8, with r+0.966)" this value is higher than ATR. These are all different values compared to the ATR document. Table 3 n + 40 and range 14-138 same as ATR.

Page 7: :The data for "38 subjects, for rate of BAC decrease with time in the postabsorptive period are summarized in Table 4, expressed as hourly clearance rates, B 60." (40 subjects ATR). "From the individual vales for B 60 and the individual curves for BAC versus time, Co values were calculated by regression and, together with the alcohol consumed, employed to calculate values for Widmark's distribution factor r for the ratio of body alcohol concentration: BAC for 38 subjects." (39 subjects ATR.)

Page 8: Table 4 "the data for 38 subjects", (39 subjects ATR). Table 4: all of the values are changed except for the N=14 females and the mean for the total 0.71. "The results for 37 subjects" (39 subjects ATR.) The values for mean are higher than ATR, S. D. are lower, C. V. values are all lower, and range values are all higher except for one value.

Page 9: Table 5 Means two values are higher and one the same; S. V. values all higher, C. V. values one higher and two are lower; Range values are mixed (052-0.95 males IMJFM with 0.60 - 0.87 males ATR; 0.52 - 0.77 IMJFM females and 0.54 - 0.85 females ATR; 0.52 - 0.95 range IMJFM and 0.54 - 0.87 ATR. All of the values for Table 6 have been changed. "Peak BAC's between 72 to 156 mg/dl. Table 6: Females 13 IMJFM and Females 14 ATR. All Means are higher than ATR, all S. D. are higher than ATR, C. V. values are higher than ATR, and the Range is as follows: Males 77 -155 IMJFM and 87 - 148 ATR; Females 98 - 156 IMJFM and Females 101 - 146 ATR; Total 77-156 IMJFM and Total 87 - 148 ATR.

Pate 10: " the mean peak BAC of 113.0 mg/dl". (131.2 mg/dl ATR.) The effect/Dose for "This Study 114" but (compares to 124 ATR). Study by Sidell and Pless 1971 is missing from the ATR document.

Page 11: "rate of BAC decrease between 10.4 and 23.8 mgl-1hr-1" , however on "Table 4 the range of rates are 10.8 to 26.2" and (rate of BAC decrease between 8.9 and 23.3 mgdl-1hr-1 in ATR on Table 4 the range was 10.6 or 22.1.) "For our data on 37 subjects" (38 subjects ATR.) The values "y= 0.041x + 12.2 and r2 = 0.147". (y + 0.32x + 12.2 and r2 + 0.093 ATR.) "For N =37, this r=0.383 value barely exceeds the critical value r 0.025=0.382" and (N = 38, the r = 0.305 value exceeds the critical value of r 0.05 = 0.05 = 0.271 ATR.) "PT=20.54 for 36 degrees of freedom at a confidence level of 99.5%" and (PT = 22.15. For 37 degrees of freedom or at a confidence level of of 99.5% ATR.) "tcalc = 20.54/>t table =2.70" and (22.15 />t table = 2.72 ATR.)

Page 13: "Table 9 B 60 is 17.1 and Widmark's R 0.74 for men and 0.65 for women". (B 60 102 for men and 104 for females and for Widmark's R 0.73 for men and 0.66 for females Table 8 in ATR.) Study Abele is omitted from the ATR document and Frasier added to the ATR document. "(T)he typical 70-kg subject" and ("typical 75-kg subject" ATR.)

Page 14: All the variables in Figure 3 have been changed on the ATR document.

Page 15: "nearly 16 hours to reach alcohol-free status". ("nearly 17 hours to reach alcohol-free status".)

Page 16: Sidel reference omitted.

This material was presented by Dubowski in ANNALS OF CLINICAL CHEMISTRY AND LABORATORY SCIENCE, (abstract). The numbers of the study are the same, however there are differences in the Peak BAC, range from the study published in ALCOHOL TECHNICAL REPORTS, but the numbers of subjects are the same as the INTERNATIONAL MICROFORM JOURNAL OF LEGAL MEDICINE. Dubowski had trouble with the number of subjects and these appear to be blood not breath concentrations.

Why were so many numbers changed? Even with the changes, there are many gross errors. Much of this should be accounted for. Many of the errors are simple mathematical ones which would/should have been found had the documents been peer reviewed.

This report makes the ATR report even more interesting for Jensen to have brought to court since they both use Widmarks r and B 60. Widmark himself stated he wasn't sure that his formula could be use on breath alcohol concentrations.

Dubowski, K. M.; "Human pharmacokinetics of ethanol: further studies," CLINICAL CHEMISTRY (1976b), 22:1199. , Dubowski states there are "78 normal male volunteers" however, on the chart under N there are "All 142". On the next column "N is 133" at the bottom. "We conclude that zero-order kinetics apparently do not apply to postabsorptive BAC decline with time, and that speculative retrograde extrapolation of BAC is unwarranted." [Why are the numbers different? We don't know if these are blood or breath alcohol concentrations or did he take the two and average them? Why are the number of subjects different in each N column? Why not break out the data by administrated amount since three levels were use? Not peer reviewed.]

"The data reported here are useful for experimental purposes and as a guarded basis for clinical judgments in acute alcoholic intoxication. However, they also demonstrate the hazards attending predictions of expectable peak BAC's after social alcohol consumption or of the time required for reduction of an initial BAC to a given lower value, as well as the limitations, in forensic practice, of speculative retrograde BAC extrapolation using an assumed universal rate of BAC decline," (Dubowski, 1976, 55-63). Not blood alcohol concentrations.

"The technically difficult problem of continuous monitoring of ethanol in expired alveolar air, further to confirm "steepling", appears not to have been undertaken to date," ( Dubowski, 1976, page 29). Dubowski didn't research this.

Yap et al. studied the chronopharmacology of ethanol and found "significantly higher peak blood ethanol concentration was attained at the 9 am session, other time-of-day differences did not reach significance and the pharmacokinetics of ethanol were essentially unchanged. It appears that major circadian variability in the metabolic and/or behavioral effects of ethanol is unlikely to occur. There was no suggestion of any differences in ethanol elimination rate in our study," (Yap, 1993, p. 21-22).

Zeiner et al. (1980) "found time of day a significant factor in peak blood ethanol concentration (BAC) in 69 Caucasian males and females. Groups received .52 ml of ethanol per kg of body weight in 5 min period during morning (08-12 hrs), afternoon 12-15 hr) and evening (15-20 hr). The peak BAC was higher and the elimination rate faster for morning drinkers than for evening drinkers, the afternoon group was intermediate."

Wilson et al conduced a study specifically directed toward the biorhythmicity of ethanol metabolism. The authors concluded that 'the rate of disappearance of alcohol does not proceed at a constant rate, but varies with the time of day in a pattern determined by the sleep habits of the individual', rather than a response to specific sleep episodes," (Sturtevant, 1989).

Sturtevant and Sturtevant (1989) believe "a common, serious shortcoming in protocol design of studies investigating variation in biological phenomena is that of too few experimental time points. It is particularly unfortunate that this flaw continues to be overlooked by investigators, reviewers, and editors alike, as noted even in recent publications. Studies from a number of laboratories have confirmed the earlier observation of temporal variation in the maximal BAC measured after a given dose of ethanol, with the highest values attained after a morning dose. It should be noted here that although several investigators, including ourselves, have found it expedient to estimate BAC values indirectly via breath ethanol readings for purposes of rapid calculation of successive doses. BAC values estimated from Breathalyzer readings are not sufficiently accurate for pharmacokinetic purposes. The method is subject to gross individual errors," (Sturtevant and Sturtevant,1989).

"Breath-alcohol analyzers are less useful than direct analysis of blood. The precision, accuracy and linearity of breath instruments are not well established at the relatively low concentrations of alcohol reached after drinking small doses of alcohol. Moreover, the apparent blood-breath ratio of ethanol changes with time after drinking and its value depends on many factors that are difficult to control in practice," Jones (1991, p. 435).

"(T)he breath-alcohol concentrations (r=-0.002 + 0.89x (BAC), which implies that the breath alcohol readings (g/210 L) are on the average 10% less than the corresponding BAC measurements (g/100 mL)," Jones (1996: 917-8).

"Wagner et al. recently proposed another way to estimate the amount of alcohol ingested from a single measurement of BAC, without having to assume values for Widmark's B and r . Their method entailed plotting the BAC measured at 2 hours post-drinking against the dose of alcohol administered in g/kg and finding the regression relationship between these variables. We have shown that result obtained by Widmark's method and the regression equation proposed by Wagner, et al. agreed almost exactly (unpublished work). For most practical purposes Widmark's equation provides a fast and reliable way to estimate the quantity of alcohol consumed provided that the limits of uncertainty are considered," (Gullberg and Jones, 1994).

Watson et al. (1981) wrote "It is nearly 50 years since Widmark first published this important equation, and while still basically correct, to use it in the same form today is to fail to take advantage of the data from numerous studies on body fluids published since 1950. We show herein that the Widmark equation can be recast, with use of these data, to yield more accurate predictions of BAC.

"The Widmark equation describes the relationship, under fasting conditions, between the total amount of alcohol ingested (A), the alcohol concentration in blood at zero time (Co) and the body weight (p): A = r x p x Co [equation 1]

"Here r is a factor defined as the fraction of the body mass in which alcohol would be present if it were distributed at concentrations equal to that in blood.

"The rate of alcohol metabolism in humans is linear under most conditions, particularly at moderate BAC's, and under such conditions, the Widmark equation can be modified by substituting the term

"Co = Ct + Bt [equation 2] so that equation 1 becomes

"A = r x p (Ct = Bt) [equation 3] where Ct is the blood alcohol concentration after time t and B is the rate of ethanol disappearance from the blood, i.e., the slope of the descending portion of the blood alcohol curve.

"Values for r can be determined if all other variables in equation 1 or 3 are known. However, in most situations in which the Widmark equation is used, such as in predicting BAC's, all other variables are not known, and r must be assigned a value. From measurements on 30 subjects (20 men), Widmark assigned to r mean values of .68 + .085 for men and .55 + .055 for women. Since these estimations in 1932, other investigators have reported values ranging from .50 to .95. In cases in which r was not measured under fasting conditions, values greater than 1.0 have been recorded, presumably due to delayed absorption of alcohol because of food in the gastrointestinal tract or alimentary tract.

"The decision on what value to ascribe to r can be entirely avoided by reducing the problem to first principles. The composition of the total body mass can be described as the sum of the body fat mass (i.e., ether-soluable lipids) and the lean body mass. The latter may be fined as the total body was mass plus the total lean solids mass. Since alcohol does not dissolve in body fat to an appreciable extent but is freely miscible with water, ingested alcohol will be almost totally associated with the body water. Accordingly, in any tissue, the alcohol content will be proportional to the tissue's water content. Alcohol is thus suitable for use as a solute to measure that tissue's body water volume and, in fact, has been used for this purpose. If we assume that when alcohol is ingested it is absorbed immediately and spreads instantaneously and uniformly throughout the body water, it follows that the concentration of alcohol in the total body water equals that in the blood water.

"If the weight (in grams) of alcohol ingested by fasting subjects is A , the total volume (in liters) of water in the body is TBW, and the hypothetical zero-time equilibrated blood alcohol concentration (in grams per liter) is Co, then it follows that A /TBW = Co/Bw and A = TBW x Co /Bw [equation 5] or using equation 2,

A = TBW /Bw x (Ct + Bt) [equation 6]

"In deriving these equations it has been assumed that elimination of alcohol from the body follows zero-order kinetics. The TBW model is based on the following equations using weight, height, and age. "TBW = 2.447 - 0.09516 Age + 0.1074 Height (cm) + 0.3362 Weight (kg). [equation 7]

"The derivation of these equations is fully reported elsewhere, together with nomograms for estimating TBW in men and women based on these equations. Equations 7 through 10 will need to be revised in the future if a major change in lifestyle results in a significant shift in the mean TBW of a population.

"The TBW of any person thus can be either measured directly, calculated using one of the above equations or read from the appropriate nomograms. The fraction of water in the blood (Bw) can also be measured directly, or a mean value of .80 can be used, depending on the accuracy required. Normal variations from this mean value are small and the error introduced by using .80 is acceptable. With this value for Bw equations 5 and 6 then become

A = TBW/.80 x Co [equation 11] and A = TBW /.80 x (Ct + Bt) [equation 12]

"The improved accuracy obtained by using the TBW estimation equations may be demonstrated if actual measured Co values are compared with those predicted by equation 11, and by the Widmark equation, transformed as follows:

Co = .80/TBW x A Co = A [equation 13] and Co = A/ r x p [equation 14]

"Hence, not only does the Widmark equation overestimate the blood alcohol level in most cases, but it also predicts Co values much less precisely, particularly in women," ( Watson, 1981).

York and Hirsch (1997) estimated the volume of distribution for ethanol in a racily mixed group of 276 alcoholics and 166 non alcoholics (aged 20-59 years) by means of bioelectric impedance methodology (BIA) and by means of prediction equations based upon age, body weight, and height. Estimations of mean TBW from BIA were found to be only slightly higher (1-4%) than those provided by the prediction equations TBW values generated from both prediction equations were also highly correlated with TBW values obtained by impedance methodology, with the highest correlations observed in females (particularly black) and in alcoholics (particularly female).

Dubowski's "Breath alcohol analysis"
There are problems with "Breath-alcohol analysis" (Dubowski, 1976: page 28). "Because of the usual interval between apprehension and the taking of a breath sample, very large arterio-venus differences are no doubt infrequent, but the circumstances of some arrests and their time frames strongly suggest that failure of the subjects to be post-absorptive could lead to his breath's yielding a significantly higher calculated blood concentration than present in venous blood."

"The technically more difficult problem of continuous monitoring of ethanol in expired alveolar air, further to confirm "steepling', appears not to have been undertaken to date. A forensic problem encountered at trial with the results of both breath and blood analyses (and apparently in some jurisdictions, analyses of urine) is that of a witness being asked to estimate the ethanol concentration which was present in the biological material concerned at a prior time considerably removed from the time of taking of the sample actually analyzed; that is to engage in speculative 'retrograde extrapolation.' If attempted, it must be based on assumptions of uncertain validity, or the answer must be given in terms of a range of possible values so wide that it is rarely of any use. If retrograde extrapolation of a blood concentration is based on a breath analysis the difficulty is compounded. "It is unusual for enough reliable information to be available in a given case to permit a meaningful and fair value to be obtained by retrograde extrapolation. If attempted, it must be based on assumptions of uncertain validity, or the answer must be given in terms of a range of possible values so wide that it is rarely of any use, (Page 29).

"The status of absorption and distribution does not affect the presumed blood concentration obtained by calculation if the value calculated is specified as that obtaining (sic) in pulmonary venous blood. Any value reported is that presumed present in the sample when obtained, and no estimate of the value at any earlier time should be attempted. Such retrograde extrapolation is not needed if the offense is defined by stature in terms of a quantity found at the time the specimen was taken with appropriate limitations on the time lag allowed.

"Data are not presently available to permit an opinion as to whether the steepling effect presents a forensic problem for breath analysis, expert in the comparison of results of analysis for ethanol in nearly simultaneous blood and breath specimens," (Page 32).

"Breath analysis for ethanol, especially in respect to forensic aspects, has been reviewed. Included are matters dealing with instrumentation, physiological factors involved in the elimination of ethanol via the breath, and especially, the uncertainties in the calculation of a whole blood concentration of ethanol from the quantity found in breath.

"Summary. We believe that the conversion of a breath quantity to a blood concentration of ethanol, for forensic purposes should be abandoned and that the offense of driving while under the influence of alcohol should be statutorily defined in terms of the concentration of ethanol found in the breath in jurisdictions employing breath analysis. The breath sample would require some extension of present federal standards for evidential breath-testing devices," (Mason and Dubowski, 1976, page 33). Also dated material, see Hlustala, 1998.

Dubowski article "Absorption, distribution and elimination of alcohol"
A few problems in "Absorption, distribution and elimination of alcohol" were discovered concerning the article aside from it not being peer reviewed. It was presented as a paper at a conference. No mention is made of peer review and there are no dates for submission and acceptance. "Absorption, distribution and elimination" Dubowski, 1985. This was published in a supplement which was not peer reviewed.

Pages 99-100 Table 1 not found in article cited. Dubowski, K. M,., "Human pharmacokinetics of ethanol: further studies," CLINICAL CHEMISTRY (1976b), 22, 1199. On page 101 paragraph 2 he states there were "134 adult men in the postabsorptive state" there were 78 men in the study, (Dubowski, 22, 1199. On Table 2 "N = 79 + 69 + 145 " [(sic)148]. Table 2 shows data which was not published in study cited, (Dubowski,1976, 22:1199).

Page 103 paragraph 3, Dubowski makes note of "circadian shifts". "Others were unable to demonstrate significant time of day-related differences in the pharmacokinetic parameters".

Page 103 paragraph 5, "The same phenomenon is illustrated in Figure 2, which consists of six representative breath alcohol versus time curves obtained in the studies of Dubowski (1976b). These figures were never published in that study--Dubowski, K. M,, "Human pharmacokinetics of ethanol: further studies, CLINICAL CHEMISTRY (1976b), 22, 1199.

Page 105 paragraph 4, Dubowski states "curves in figures 2A -F . . . show simultaneous blood alcohol and breath alcohol concentrations in the course of the experiments that yielded the curves in Figures 2A-F from Dubowski, KM and O'Neill, B, "The blood/breath ratio, of ethanol, CLINICAL CHEMISTRY (1976), 25 (6):1144. The graphs are marked Breath-alcohol--all of them on page 104. [Where are the blood alcohol graphs? They (the six graphs) don't appear in the abstract from CLINICAL CHEMISTRY (1976), 25 (6):1144 as cited.]

Page 106 paragraph 3, "Breath alcohol time curves are subject to . . .Fifth significantly large short-term fluctuations occur in some subjects and results in marked positive and negative departures from the alcohol concentration trend line. Sixth, short-term marked oscillation of the blood or breath alcohol concentration can occur at various points of the curve, resulting in repeated excursion of the alcohol concentration above and below a given concentration within a few minutes or hours."

Research on alcohol and curves
Fifty percent of the alcohol-related crashes occurred on curves compared with only 36% of the 'sober' crashes. In negotiating a curve, two behaviors are critical: an accurate estimation of the degree of curvature to enable appropriate speed selection and precise tracking of the curve path. It is a conventional divided attention-task and alcohol is known to seriously impair timesharing ability. Making the reasonable assumption that impaired drivers will devote relatively more attention to their short-term steering task it is likely to be their perception of the relevant cues to curvature that will suffer, resulting in in an inappropriate curve entry speed which in turn will exacerbate the demands placed on steering control, (Johnston, 1982).


al-Lanqawi, Y.; Moreland, T. A.; McEwen, J.; Halliday, F.; Durin, C. J.; Stevenson, I. H.; "Ethanol kinetics: extent of error in back extrapolation procedures," BRITISH JOURNAL OF CLINICAL PHARMACOLOGY (1992), 34 (4) : 316-321.

Baselt, R. C.; Danhof, I. E.; "Disposition of ethanol in man"; In: J. C. Garriott, ed., MEDICOLEGAL ASPECTS OF ALCOHOL DETERMINATION IN BIOLOGICAL SPECIMENS : volume 2, PSG Company, Inc., 1988, p. 59.

Baselt, R. C.; "Disposition of ethanol in man"; In: J. C. Garriott, ed., MEDICOLEGAL ASPECTS OF ALCOHOL DETERMINATION IN BIOLOGICAL SPECIMENS : volume 3, Lawyers and Judges Publishing Company, Inc., 1996, p. 66.

Cooper, W. E.; Schwar, T. G.; Smith, L. S.; Alcohol, drugs and road traffic; Juta & Company Limited, c1979, 402p.

Dubowski, Kurt M.; "Absorption, distribution and elimination of alcohol: highway safety aspects," JOURNAL OF STUDIES ON ALCOHOL (1985) 10 (Supplement) : 98-108.

Dubowski, K. M.; and O'Neill, B.; "The blood/breath ratio, of ethanol," CLINICAL CHEMISTRY (1976), 25 (6):1144.

Dubowski, KM, "Human pharmacokinetics of ethanol," INTERNATIONAL MICROFORM JOURNAL OF LEGAL MEDICINE (1975), 10 (4): 1-16.

Dubowski, Kurt M.; "Human pharmacokinetics of ethanol I : peak concentrations and elimination in male and female subjects," ALCOHOL TECHNICAL REPORTS (1976), 5 (4) : 55-63.

Dubowski, K. M.; "Human pharmacokinetics of ethanol: further studies," CLINICAL CHEMISTRY (1976b), 22:1199.

Dubowski, K. M.; "The role of the scientist in litigation involving drug-use testing," CLINICAL CHEMISTRY (1988), 34 (4): 788-792.

Dubowski, K. M.; "Studies on pharmacokinetics of ethanol in man," ANNALS OF CLINICAL CHEMISTRY AND LABORATORY SCIENCE (1975), 5 (4): 309-10.

Fisher, Honey R.; Simpson, Robert I.; Bhushan, M. Kapur; "Calculation of blood alcohol concentration (BAC) by sex, weight, number of drinks and time," CANADIAN JOURNAL OF PUBLIC HEALTH (1987), 78, September/October : 300-304.

Gengo, R. M.: Gagos, C.; Straley, C.; Manning, C.; "Pharmacodynamics of ethanol: effects of performance and judgment," JOURNAL OF CLINICAL PHARMACOLOGY (1990), 30 (8): 748-754.

Gullberg, R. G.; "Employing simulated date to illustrate an important cause of the 'steepling' effect in breath alcohol analysis," MEDICINE, SCIENCE, AND THE LAW (1994), 34 (4): 321-3.

Gullberg, R. G.; Jones A. W.; "Guidelines for estimating the amount of alcohol consumed from a single measurement of blood alcohol concentration: Reevaluation of Widmark's equation," FORENSIC SCIENCE INTERNATIONAL (1994), 69 (2) :119-130.

Gullberg, R. G.; "Reproducibility of within-subject breath alcohol analysis," MEDICINE, SCIENCE, AND THE LAW (1998), 38 (2): 157-62.

Gullberg, R. G.; "Variation in blood alcohol concentration following the last drink," JOURNAL OF POLICE SCIENCE AND ADMINISTRATION (1982), 10 (3) :289-296.

Hlastala, M. P.; "The alcohol breath test--a review," JOURNAL OF APPLIED PHYSIOLOGY (1998), 84 (2): 401-408.

Jaffe, J. H.; Encyclopedia of Drugs and Alcohol, New York, Macmillan Library Reference USA, 1995, 4 vols. Articles on Alcohol, BAC, Drug Metabolism, Drunk Driving, Pharmacokinetics of Alcohol, etc.

Jakus, J. T.; Shajani, N. K.; and Image, B. A.; "Consumption of a large dose of alcohol in a short time span," FORENSIC SCIENCE INTERNATIONAL (1992), 56: 113-125.

Jones, A. W.; "Biochemistry and physiology of alcohol: applications for forensic science and toxicology" ; In: J. C. Garriott, ed., MEDICOLEGAL ASPECTS OF ALCOHOL DETERMINATION IN BIOLOGICAL SPECIMENS : volume 3, Lawyers and Judges Publishing Company, Inc., 1996, p. 99.

Jones, A. W.; "Concentration-time profiles of ethanol in capillary blood after ingestion of beer," JOURNAL OF FORENSIC SCIENCES SOCIETY (1991), 31 (4): 429-439.

Jones, A. W.; "Disappearance rate of ethanol from the blood of human subjects: implications for forensic toxicology," JOURNAL OF FORENSIC SCIENCE (1993), 38 (1) : 104-118

Jones, A. W.; "DUI defenses" ; In: Steven B. Karch, Ed.; DRUG ABUSE HANDBOOK, CRC Press, 1998, 1138p.

Jones, A. W.; "Enforcement of drink-driving laws by use of 'per se' legal alcohol limits: blood and/or breath concentrations as evidence of impairment," ALCOHOL, DRUGS AND DRIVING ( 1988), 4 (2): 99-112.

Jones, A. W.; Jonsson, K. A.; "Food-induced lowering of blood-ethanol profiles and increased rate of elimination immediately after a meal," JOURNAL OF FORENSIC SCIENCE (1994), 39 (4) : 1084-93.

Jones, A. W.; "Inter individual variation in the disposition and metabolism of ethanol in healthy men," ALCOHOL (1984), 1 (5): 385-91.

Jones, A. W.; Jonsson, K. A.; Neri, A.; "Peak blood-ethanol concentration and the time of its occurrence after rapid drinking on an empty stomach," JOURNAL OF FORENSIC SCIENCE (1991), 36 (20) : 376-385.

Jones, A. W.; "Pharmacokinetics of Alcohol"; In: Jaffe, J. H.; Encyclopedia of Drugs and Alcohol, New York, Macmillan Library Reference USA, 1995, p805.

Jones, A. W.; Jorfeldt, L.; Hjertbert, H.; Jonsson, K. A.; "Physiological variations in blood ethanol measurements during the post-absorptive state," JOURNAL OF THE FORENSIC SCIENCES SOCIETY (1990), 30 ( 5) : 273-83.

Jones, A, W.; "Status of alcohol absorption among drinking drivers," JOURNAL OF ANALYTICAL TOXICOLOGY (1990), May-June 14 (3) : 198-200.

Jones, A. W.; "Widmark's equation: determining amounts of alcohol consumed from blood alcohol concentration," DWI JOURNAL (1986), p. 8-11.

Jones, A. W.; Andersson, Lars; "Variability of the blood/breath alcohol ratio in drinking drivers," JOURNAL OF FORENSIC SCIENCE (1996), 41 (6): 916-921.

Johnston, I. R.; "The Role of alcohol in road crashes," ERGONOMICS (1982), 25: 941-946.

Khanna, Jatinder M.; LeBlanc, A. Eugene; Mayer, Joel A.; "Alcohol pharmacokinetics and forensic issues: a commentary", In: K. E. Crow, R. D. Batt (Eds.); Human metabolism of alcohol, volume I: Pharmacokinetics, medicolegal aspects and general interest; CRC Press (1989), 214p. (59-70).

Labianca, D. A.; "Estimation of blood-alcohol concentration," JOURNAL OF CHEMICAL EDUCATION (1992), 69 (8): 628-632.

Lands, W. E. M.; "A Review of alcohol clearance in humans," ALCOHOL (1998), 15, (2): 147-160.

Mason, H.; and Dubowski, K. M.; "Breath-alcohol analysis: uses, methods and some forensic problems-review and opinion," JOURNAL OF FORENSIC SCIENCES (1976), 21(9):9-41.

Moskowitz, Herbert; "Computing a BAC estimate," National Highway Traffic Safety Administration, U. S. Department of Transportation, October 1994, 3p.

Montgomery, M. R.; Reason, M. J.; "Retrograde extrapolation of blood alcohol data: an applied approach," JOURNAL OF ANALYTICAL TOXICOLOGY (1992), 36 : 281-292.

O'Neill, B.; Williams, A. F.; Dubowski, K. M.; "Variability in blood alcohol concentrations. Implications for estimating individual results," JOURNAL OF STUDIES ON ALCOHOL (1983), 44 (2) : 222-230.

Santamaria, J. N., Ethanol ingestion studies, Department of Community Medicine, St. Vincent's Hospital, Fitzroy, Australia Department of Transport, Office of Road Safety, 1979.

Shajani N. K.; Dinn, H. M.; "Blood alcohol concentrations reached in human subjects after consumption of alcoholic beverages in a social situation," CANADIAN SOCIETY OF FORENSIC SCIENCES JOURNAL (1985), 18 (1): 38-48.

Sturtevant, Ruthann P; Sturtevant, Frank M.; "Circadian variation in rates of ethanol metabolism"; In: K. E. Crow, R. D. Batt (Eds.); Human metabolism of alcohol, volume I: Pharmacokinetics, medicolegal aspects and general interest; CRC Press (1989), 214p. (23-39).

Thompson, Robert Q.; "The Thermodynamics of drunk driving", JOURNAL OF CHEMICAL EDUCATION (1997), May,.

Wagner, John G.; Wilkinson, Paul K.; Ganes, Derek A; "Estimation of the amount of alcohol ingested from a single blood alcohol concentration," ALCOHOL AND ALCOHOLISM (1990), 25 (4) : 379-384.

Watson, Patricia E. Watson, Ian D.; Batt, Richard D.; "Prediction of blood alcohol concentrations in human subjects: updating the Widmark equation," JOURNAL OF STUDIES ON ALCOHOL (1981), 42 (7) : 547-556.

Watson, Patricia E.; "Total body water and blood alcohol levels: updating the fundaments," In: K. E. Crow, R. D. Batt (Eds.); Human metabolism of alcohol, volume I: Pharmacokinetics, medicolegal aspects and general interest; CRC Press (1989), 214p.

Widmark, E. M. P.; 1932, Principles and Applications of Medicolegal Alcohol Determination, translated into English by R. C. Baselt, Department of Pathology, University of California, Davis, 1981, 163 p.

Winek, C. L.; Esposito, F. M.; "Blood alcohol concentrations: factors affecting predictions," LEGAL ISSUES IN MEDICINE (1985) : 34-61.

Yap, Madeline; Mascourd, D. J.; Starmer, G. A.; Whitfield, J. B.; "Studies on the chronopharmacology of ethanol," ALCOHOL AND ALCOHOLISM (1993), 28 (1): 17-24.

York, J. L.; Hirsch, J. A.; "Applications of bioelectric impedance methodology and prediction equations to determine the volume of distribution for ethanol," ALCOHOL, 12 (6) : 553-8.

Zeiner, A.; Andress, D.; Stratton, R.; "Time of day effects on peak blood alcohol concentration and clearance rate" (Abstract), ALCOHOLISM AND CLINICAL EXPERIENCES IN RESEARCH (1980), 4 : 233.

Zink, P.; Reinhardt, G.; "Course of the blood alcohol-curve during and after consumption of alcohol in large quantities: long time studies on human volunteers. . .In: K. Kuriyama; A. Takada; H Ishii (Eds.) Biomedical and social aspects of alcohol and alcoholism, Amsterdam, The Netherlands: Elsevier Science Publishers B. V., 1988, 892p. (p.623-628).


updated 12/16/16