Variability of Ethanol Absorption and Breath Concentrations During a ...

Vol. 19, No. 4 August 1995

0145-6008/95/1904-1055$03.00/0

ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH

Variability of Ethanol Absorption and Breath Concentrations During a Large-Scale Alcohol Administration Study Patrick N. Friel, John S. Baer, and Barry K, Logan

Ethanol disposition was evaluated in 77 female and 97 male college seniors during an alcohol challenge study. All were regular drinkers who exceeded legal intoxication levels at least twice a month by history. A standard ethanol dose (females, 0.43 glkg; males, 0.51 g/kg) was administered over 10 min, after a 4-hr fast, and breath alcohol concentrations (BrACs) were measured for 2 hr. lntersubject variability in BrACs was greatest early in the study, during ethanol absorption; the coefficient of variation decreased from 39% at 14 rnin to 14% at 125 min after the start of drinking. The time to peak BrAC varied from 10 to 91 min after the start of drinking (mean 39.6 min). Mean BrACs were significantly lower in females than males; mean peak BrACs were 0.054 g/210 liters in females and 0.058 g/210 liters in males ( p = 0.031). The p- and r-values for both genders were higher than those typically used in ethanol dose calculation formulas. Data are discussed to direct future research. The constants used in Widmark’s formula need to be revised differentially for males and females in this population to reach specific target BrACs. Furthermore, substantial variability in absorption rates must be accounted for when assessing rising versus falling limb BrAC phenomena. Key Words: Ethanol, Absorption, Breath Alcohol, Widmark, Gender.

in blood (r) describes ethanol distribution, and the postabsorptive linear rate of decay in ethanol concentration (p) is a measure of elimination (primarily via metabolism). Ethanol’s absorption rate is known to be a function of several factors, including dose and gastric emptying time, which is influenced by whether alcohol is consumed on an empty or full stomach.’ After absorption, ethanol is distributed into body water; r is proportional to relative body water content, and typically higher in men than women. Ethanol elimination is less variable between subjects than absorption, but the elimination rate tends to increase with heavy drinking.8,10Most alcohol research is done in fasting subjects to minimize intersubject variability in absorption, but some variation in time-to-peak BrAC is still encountered. We conducted this study to characterize ethanol absorption and the intersubject variability in BrACs during a largescale alcohol administration study involving a well-defined sample of both males and females. We also determined how closely subjects’ BrACs apINGLE-DOSE alcohol challenges, followed by detailed proached target concentrations, which were based on Widpsychophysiological testing, are widely used in alcohol- mark’s The analyses are designed to assist in ism research. For example, sensitivity to alcohol effects is the development of future studies and also provide one of one hypothesized risk factor for the development of alco- the few large-scale pharmacokinetic data sets for ethanol in hol-related Several recent publications have women. discussed the importance of intersubject variability in blood or breath alcohol concentrations (BrACs) in these studies. METHODS Variability in alcohol concentration should contribute to variability in physiological and psychological responses; Data used in this study were collected as a part of a larger longitudinal some alcohol effects also depend on whether its concentra- study conducted at the University of Washington, which is directed by the second author. The study was designed to evaluate and follow-up a large tion is rising or f a ~ l i n g . ~ - ~ sample of college seniors who drink heavily. Subjects entering the study The pharmacokinetics of ethanol are well character- were 21-25 years old and reported a pattern of drinking enough alcohol to ized.6-s Three key parameters can be used to describe the reach an estimated blood ethanol level of at least 0.12 d100 ml at least time course of BrACs after drinking. Time-to-peak concen- twice a month. Eighty-eight percent of the subjects were Caucasian. In the tration in blood or breath (Tma) is an index of absorption initial laboratory phase of the study, subjects were given a standard rate, the ratio of ethanol concentration in the body to that ethanol dose (males, 0.51 gikg; females, 0.43 g/kg ethanol; 100-proof

S

From Washington State Toxicology Laboratory (P.N.F., B.K.L.); and the Department of Psychology (J.S.B.), College ofArts and Sciences, University of Washington, Seattle, Washington. Received for publication October 17, 1994; accepted March 2, 1995 This study was supported by the National Institute on Alcohol Abuse and Alcoholism Grant AA08632 (to J.S.B.). Reprint requests: Pam‘ck N. Friel, B.S., Department of Laboratory Medicine, University of Washington, Park 901.5, Suite 360, 2203 Airport Way South, Seattle, WA 98134-2027. Copyright 0 1995 by The Research Society on Alcoholism. Alcohol Clin Erp Res, Vol19, No 4, 1995: pp 1055-1060

vodka diluted 1:5 with carbonated mixer). Drinking laboratories were conducted between 2:OO and 8:OO PM. Subjects were instructed to fast for 4 hr before the laboratory, but to eat a normal breakfast (or early lunch for later laboratories). Upon entry to the laboratory, subjects completed a form labeled “Affidavit” and “Recent Dietary Record,” noting all consumption over the past 24 hr. Sixteen of 174 subjects reported eating within the 4-hr fasting period, but none within 2 hr of the time of the laboratory. The doses were calculated based on Widmark’s formula, and were expected to give theoretical peak BrACs (Co) of -0.075 g/210 liters, and 1 hr BrACs of -0.060 g/210 liters.”.” Subjects were given 10 min to drink the alcohol, and rinsed their mouths with water after drinking and before giving the first BrAC, to

1055

FRlEL ET AL.

1056 Table 1. Mean Breath Alcohol Data (g/210 liters) During an Alcohol Challenge Study ~~~

Sample time (min)

Females (n)

0

0 (77) 0.042(66) 0.049 (77) 0.047(77) 0.046 (77) 0.044(77) 0.033(72) 0.030(46)

14 29 46 59 72 105 125

Males (n)

p (t test)’

-

0

(97) 0.047 (88) 0.054(97) 0.052(97) 0.051 (97) 0.048(97) 0.036 (91) 0.032(73)

0.097 0.014 0.003 0.000 0.000 0.006 0.006

All subjects (n) 0 (174) 0.045(154) 0.052(174) 0.050(174) 0.049 (174) 0.046 (174) 0.035(163) 0.031 (119)

cv (%)

39 27 20 17 17 18 14

’Two-tailed unpaired t test with null hypothesis of equal means for females and males.

eliminate as much residual mouth alcohol as possible. During the laboratory, subjects were seated alone at a desk with a computer terminal in a small room. They were kept mentally active by computer-based questions and by performing BrAC tests. Stress stimuli, consistingof the anticipation (via auditory tone) and experience of mild electric shocks, were administered at 15 rnin before and 20 and 65 min after the start of drinking. BrAC readings were collected at -15, 30, 45, 60, and 75 rnin after the start of drinking. Additional readings were taken if needed at 105 and 125 min, until the BrAC was 0.03 gI210 liters, when the subjects were discharged from the laboratory. Duplicate readings were collected for each of the 15to 75-min samples; thereafter, only single BrAC readings were available for most subjects. BrACs were collected using the BAC Verifier Datamaster I1 (National Patent Analytical Systems, Inc., Mansfield, OH), which was calibrated before the study using an alcohol-water simulator solution providing vapor ethanol concentration of 0.10 g/210 liters at 34°C. Before the study, a quality assurance check verified that imprecision was 53% and inaccuracy was 55%. This study was conducted with the data from the first 185 subjects to complete this phase of the study. Eleven subjects were excluded because of missing data, leaving a sample of 174 subjects (77 females and 97 males). They consumed alcohol on a median 3 dayshveek (range 1-7), with a median of 5 standard drinks/ocapsion. Forty-four females took oral contraceptives. Data concerning menstrual cycle phase were not collected. Because the first BrAC was taken -5 rnin after the end of drinking, initial duplicate breath test results were evaluated carefully for evidence of mouth alcohol. In any case in which mouth alcohol was suspected to have contributed to the first reading, that reading was excluded from calculations. Criteria for possible mouth alcohol were duplicates not in agreement by forensic standards (? 10% of the mean), or a result inconsistent with the remaining data. The initial reading was excluded in 20 of 174 cases, in which the earliest possible time-to-peak BrAC became the time of the second sample (-30 rnin after the start of drinking). Peak BrAC (CmJ and time-to-peak BrAC (Tma) were determined for each study subject. Pharmacokineticmodeling used Minim 1.8a for Macintosh computers (Dr. R. D. Purves, Department of Pharmacology, Medical School, University of Otago, Dunedin, New Zealand). Data were fit to a one-compartment open model with first-order absorption (rate constant k) and zero-order elimination (rate constant 01, using the following equation: Ct

=

CO(I - e-kt) - (Pt) .

liters by the 105-min sample (Table l), we used mean data only out to 105 min for modeling, and for Fig. 1 (A and B) to avoid bias.

RESULTS

BrAC Reproducibility

Mean BrAC data at each data collection time are given in Table 1, for males, females, and the combined study group (the times given are the actual mean sampling times). The coefficient of variation (CV; SD/mean) was calculated for the combined BrACs at each time point. The CV is a useful measure of the variability of a measurement at different concentrations. The CV declined steadily from 39% at the first (14-min) sample time to 14% at the last (125-min) sample. A visual display of the variability in BrACs over time is given in Fig. 1 (A and B). Subjects were grouped according to their time-to-peak BrAC, and then mean BrAC values were calculated at each time point for each group. Mean BrACs varied between groups by >2-fold at the 14-min sample; but, by 1 hr after the start of drinking, the curves were nearly superimposable. Gender Difperences

Mean female BrACs were consistently lower than mean male BrACs, and the difference was statistically significant after the second BrAC (Table 1). Thus, the ethanol doses for males (0.51 gkg) and females (0.43 gkg) used in this study did not result in equal BrACs for the two groups. Mean BrAC results were not statistically significantly different for women taking oral contraceptives, when compared with women who do not. Ethanol Absorption

The mean time-to-peak BrAC for all subjects was 39.6 2 19.1 min (mean It_ SD) from the start of drinking (range 10-91 min). Mean times to peak for males and females did not differ at the 0.05 level of significance. All of the subjects r = Ethanol Dose (gPlrg)/[Co(g/lOO g) X 101. completed their beverage, and none vomited. Although Computer fits of individual BrAC profiles gave high residuals in some subjects were given 10 min to drink, actual drinking time cases, because only limited data were available for the elimination phase varied from 4 to 13 min, with a median of 10 min, but actual of the curve. As a result, we used mean data, grouped by time-to-peak drinking time and time-to-peak BrAC were not correlated BrAC, for computer fitting. With this method, satisfactory fits to the ~ 0.05). The mean peak BrAC for all subjects model given herein were achieved with six groups of subjects (females and (r = 0 . 0 8 , > males with peak BrACs at the 14-, 29-, and 46-min BrACs). Because was 0.057 g/210 liters (range 0.033-0.126). The mean peak nearly one-third of the subjects reached the study exit BrAC of 0.03 gI210 BrAC for females (0.054 g/210 liters) and males (0.058

Co (g/210 liters) was assumed to be an unbiased estimate of Co in blood (g/lOO 9). for the purposes of calculating Widmark’s factor r with the following equation:

VARlABiLllY OF ETHANOL ABSORPTION

A

1057

MEAN BrAC‘S BY TIME TO PEAK FOR FEMALES

0.06

-

0.05

-

2 0

KJ 0.04-

3 0

2

0.03-

-0-

Peak@ 14 min. (n=13)

-0---

Peak @ 29 min. (n=27)

----Q---

Peak @ 46 min. (n-11)

--&-

Peak @ 59 min. (n-15)

0.02-

Fig. 1. (A) Mean breath alcohol readings for females in the study, sorted by time-to-peak BrAC. (B) Mean breath alcohol readings for males in the study, sorted by timeto-peak BrAC. One male with a time to peak of 91 min from the start of drinking was excluded from the figure, for the sake of clarity.

TIME (min.)

B 0.07

,

MEAN BrAC‘S BY TIME TO PEAK FOR MALES

1

* Peak @ 14 min. (n=17) --9-

Peak @ 29 min. (n-44)

----Q--

Peak @ 46 min. (n=l 1)

----I+-Peak @ 59 min. (n=14)

--*-Peak@ 72 min. (n=lO)

- T

0

25

50

75

I

I

100

125

TIME (min.) Table 2. Summary Mean Pharmacokinetic Parameters

g/210 liters) were significantly different ( p = 0.031, unpaired t test) (Table 2). Table 2 also gives the results of computer fitting of mean BrAC data to a one-compartment open model with zero-order elimination, for female and male subjects peaking at 14, 29, and 46 min after the start of the 10-min drinking period. The mean ethanol elimination rate ( p ) for females was 0.019 g/210 litershr, and for males was 0.02 g/210 litershr. Mean values for r were 0.65 for females and 0.71 for males. We found no significant correlations bemeen postabsorptive B r A ~ sand subject’s recent drinking frequency or quantity Of ethanol consumed. DISCUSSION

Our analysis of BrAC-time data has several implications for future research. First, there was substantial variability in

,C

(g/210 liters)’ T , (min)’,t ~ r3, (4/210 liters/hr)S Co (g/210 liters)*

rs

Females

Males

0.054 i 0.014 41 ? 19 0.019 0.066 0.65

0.058 f 0.010 38 f 19 0.020 0.072 0.71

’, C and T,. are given as mean 2 SD. T,. is calculated from the beginning of the drinking t 17.2% of subjects reached their peak BrAC at the 14-min sample, 40.8% at the 29-min sample, 12.6% at the 46-min sample. 16.7% at the 59-min sample, 12.1% at the 72-min sample. and 0.6% at the 91-mh sample. $ p, Co, and r are the weighted means for the parameters determined by computer fitting the mean BrAC time profiles for subjects with times to peak of 14, 29. and 46 min.

ethanol absorption (time-to-peak BrAC varied from 10 to 91 min after the start of drinking), despite the fact that ethanol was administered after a 4-hr fast and under highly

FRlEL ET AL.

1058

a reduction in the extent of ethanol absorption, an increase in its rate of elimination, or both. -‘”- I I A recent study by Breslin et a1.16 examined the effects of stress on the BrAC-time curve in 60 male and female subjects given 0.7 g/kg ethanol after a light meal. They found that subjects who watched a stressful film had increased times to peak BrAC when compared with controls. Subjects experiencing a cold pressor stress had a transient increase in postabsorptive ethanol elimination rate when compared with controls. Stressed subjects did not differ from controls in peak BrAC or area under the BrAC-time curve. They also found a significant gender effect on BrAC, even after controlling for body composition, which they attributed to a greater first-pass metabolism in males. 0.024 Our study cannot be compared directly with Breslin and coworkers’, because we administered a lower ethanol dose 0.01 0 25 50 15 100 125 under fasting conditions and used a different stress procedure. However, it is of interest that the mean initial postTIME (min.) absorptive p-values for Breslin and coworkers’ stressed Fig. 2. Comparison of theoretical and actual BrACs achieved in the study subjects (0.018 g/210 liters/hr for cold pressor and 0.017 Fable 1). Theoretical (- - -) BrACs were calculated by assuming instantaneous litershr for film stressor) agree quite closely,withour g/210 absorption of the ethanol dose at time 0, Widmark’s r of 0.68 for males (0) and 0.57 for females (0),and p = 0.015 9/21 0 liters/hr. Mean actual postabsorptive means of 0.020 g/210 litershr for stressed males and 0.019 BrACs are lower than theoretical BrACs, and female BrACs are lower than male g/210 litershr for stressed females, whereas their control BrACs. subjects’ mean p was -0.015 g/210 liters/hr. This raises the possibility that even relatively mild stress may alter ethanol controlled conditions. Variability in absorption resulted in metabolism. Previous research has shown that people with increased intersubject variability in BrACs, especially dur- serious injuries may eliminate ethanol very ra~id1y.l~ ing the first half hour of the study. Second, the actual Our data are presented in terms of BrAC, without atBrACs achieved during the study were lower than expected, tempting to convert them into blood alcohol concentration based on Widmark‘s formula (Fig. 2). Third, mean female units. This is representative of contemporary research and BrACs were consistently lower than mean male BrACs, forensic practice, wherein BrAC is often the method of even though ethanol doses were adjusted for gender using choice for determining the acute body burden of alcohol. Widmark’s factor, Y (Fig. 2). However, for some of the calculations in Table 2, we used The results of this study naturally are most easily gener- BrAC (g/210 liters), as an unbiased estimate of blood alalized to studies using similar ethanol doses in similar cohol concentration in dl00 g. Although a large body of populations and study designs. Certainly the time-to-peak research supports this assumption, small biases between data are dose-specific; lower doses will result in shorter blood and breath alcohol concentrations may occur (e.g., as times to peak BrAC, and higher doses will result in longer a result of arteriovenous blood alcohol concentration gratimes to peak BrAC.6 Ethanol administration on a full dients). Jones” found that BrAC (g/210 liters) underestistomach would also be expected to increase the time-to- mated blood alcohol (g/lOO ml) by an average of 4.2% in a peak BrAC. One important question is whether the stress pharmacokinetic study in 21 men. Blood alcohol in g/lOO g procedures themselves altered ethanol pharmacokinetics. underestimates blood alcohol in g/100 ml by -6%, because Our measure of the rate of ethanol absorption-time-tothe specific gravity of blood is -1.055. Thus, BrAC in g/210 peak BrAC-agrees closely with results from other studies liters should be a reasonable estimate of blood alcohol in in nonstressed subjects, in terms of the mean value and g/100 g. variability. For example, Dubowski13gives a mean time to This study has several implications for research design. peak blood alcohol concentration of 29.3 17.1 min after Intersubject variability in time-to-peak BrAC complicates the end of drinking in 79 males ingesting 0.5 g/kg ethanol. the assessment of the effects of rising versus falling BrACs. Similar estimates of the mean and variability of time-to- According to Newlin and T h o m ~ o n ,“[Tlhe ~ rising blood peak BrAC under fasting conditions have been report- alcohol curve is represented by approximately the first 30 ed.14,15Comparison of our time-to-peak BrAC (29.6 5 19.1 min. after drinking [depending on dose and rate of drinkmin from the end of drinking) with these data from the ing], and the falling blood alcohol curve is reflected by the literature suggests that the stress procedures used in this remaining time.” Although this is a useful generalization, study had little or no effect on the rate of ethanol absorp- only 41% of our subjects actually followed this pattern tion. However, the lower-than-expected BrACs in this (Table 2). As shown in Table 2, 17% of subjects had study might be caused by a stress effect resulting in either reached their apparent peak BrAC by 14 min after the start THEORETICAL VS. ACTUAL MEAN BrACS

I

*

VARIABILITY OF ETHANOL ABSORPTION

1059

of drinking; their BrACs were falling at 29 min (Fig. 1, A and B). At the other end of the absorption continuum, 42% of subjects had not reached their peak BrAC at 29 min after the start of drinking, and 29% of subjects had not reached their peak even by 46 min. For the researcher to be certain that BrACs are rising in all subjects, our data suggest that assessments should be conducted within a very short time after drinking; and to be certain that BrACs are falling, assessments should be conducted >60 min after drinking. (These specific guidelines naturally apply only for the ethanol dose and protocol used in this study.) Of course, a more individualized approach, assessing both rising and falling BrACs, for each subject is possible. However, frequent BrAC tests are necessary to characterize individual BrAC-time profiles. If it is necessary to reach a specific target postabsorptive BrAC in the population we studied, Widmark's constants should be revised. Our empirical estimate for Widmark's r in men (0.71) is close to Widmark's original estimate (0.68), but our r-value for women (0.65) is much higher than Widmark's (0.55).11 In other words, the apparent volume of distribution of ethanol in the women we studied was considerably higher than expected, explaining the lower BrACs achieved in women, compared with men. This may reflect a lower body fat content, and higher body water content in the young women in this study, compared with Widmark's original sample.'' In a previous study examining the reliability of Widmark's equation, using data from 115 of the subjects in this study, we found that calculated doses significantly underestimated actual doses." The present study confirms our suspicion that the bias we found was caused by bias in Widmark's constants when applied to this population and study design. The relatively rapid ethanol elimination rate ( p 0.02 g/210 litershr) in this study places the subjects at a point intermediate between social drinkers and very heavy drinkers. For example, Winek and Murphy" found a mean p of 0.012 g/210 litershr in nondrinkers, 0.015 g/210 litershr in social drinkers, and 0.030 g/210 litershr in alcoholics. An extensive review by Jones' gives a mean p of 0.016 g/100 mVhr in blood in healthy male moderate drinkers. There is some evidence that women eliminate ethanol slightly faster than but mean p-values were not significantly different in males and females in this study. However, we did not control for menstrual cycle phase. In contrast to earlier work by Jones and Jones,14 we found no evidence for inhibition of ethanol metabolism in women taking oral contraceptives. The higher p-values in this study may reflect induced hepatic microsomal enzyme activity, which normally functions as a secondary pathway for ethanol metabolism, but is thought to be responsible for accelerated ethanol metabolism in heavy drinkers? Pharmacokinetic analysis of data from a large-scale alcohol administration study in a well-defined sample of young, regular drinkers yielded several useful insights. Variability in ethanol absorption needs to be appreciated

-

for a rigorous assessment of rising versus falling limb BrAC phenomena. The apparent volume of distribution of ethanol, as measured by Widmark's r, was higher than anticipated, especially in females; subjects also eliminated ethanol at a higher rate ( p ) than expected. As a result, BrACs were lower than anticipated, especially in the case of females. Further research is needed to determine to what extent these findings can be generalized to different experimental designs. The effect of different stress procedures on ethanol's elimination rate needs to be clarified. The relatively rapid ethanol elimination rate in this study could conceivably reflect the influence of recent heavy drinking, the stress procedures, or both. In addition, ethanol absorption should be characterized under different experimental conditions (varying dose, rate of administration, stressed versus nonstressed, and fasting versus fed subjects) to determine optimal study design parameters with respect to BrAC reproducibility. ACKNOWLEDGMENTS

We are grateful to Andrew Hummel-Schluger and Sandra DiStefano for data management. Dr. Danny Shen provided a copy of Minim 1.8a software and advice on pharmacokinetic analyses. REFERENCES 1. Schuckit MA: Low level of response to alcohol as a predictor of future alcoholism. Am J Psychiatry 151:184-191, 1994 2. Pollock VE:Meta analysis of subjectivesensitivity to alcohol in sons of alcoholics. Am J Psychiatry 149:1534-1538, 1992 3. Newlin D, Thomson J: Alcohol challenge in the sons of alcoholics: A critical review and analysis. Psycho1 Bull 108:383-402, 1990 4. Nicholson ME, Wang M, Airhihenbuwa CO, Mahoney BS, Maney DW: Predicting alcohol impairment: Perceived intoxication versus BAC. Alcohol Clin Exp Res 16:747-750, 1992 5. Martin CM, Earleywine M, Musty RE, Perrine MW, Swift RM: Development and validation of the biphasic alcohol effects scale. Alcohol Clin Exp Res 17:140-146, 1993 6. Wilkinson P K Pharmacokinetics of ethanol: A review. Alcohol Clin Exp Res 4:6-21, 1980 7. Holford NHG: Clinical pharmacokinetics of ethanol. Clin Pharmacokinet 13:273-292, 1987 8. Jones AW: Disappearance rate of ethanol from the blood of human subjects: Implications in forensic toxicology. J Forensic Sci 38:104-118, 1993 9. Jones AW, Jonsson KA: Food-induced lowering of blood-ethanol profiles and increased rate of elimination immediately after a meal. J Forensic Sci 39:1084-1093, 1994 10. Winek CL, Murphy KL: The rate and kinetic order of ethanol elimination. Forensic Sci Int 27:159-166, 1984 11. Widmark EMP: Principles and Applications of Medicolegal Alcohol Determination (1932). Davis, CA, Biomedical Publications, 1981 12. Fisher HA, Simpson RL, Kapur BM: Calculation of blood alcohol concentration (BAC) by sex, weight, number of drinks and time. Can J Public Health 78:300-304, 1987 13. Dubowski KM: Absorption, distribution, and elimination of alcohol: Highway safety aspects. J Stud Alcohol 46(Suppl. 10):98-108, 1985 14. Jones MK, Jones BM: Ethanol metabolism in women taking oral contraceptives. Alcohol Clin Exp Res 8:24-28, 1984 15. Jones A W Pharmacokineticsof ethanol in saliva: Comparisonwith blood and breath alcohol profiles, subjective feelings of intoxication, and diminished performance. Clin Chem 39:1837-1844, 1993

1060

16. Breslin CF, Hayward M, Baum A Effect of stress on perceived intoxication and the blood alcohol curve in men and women. Health Psycho1 13:479-487, 1994 17. Wigmore JG, Mammoliti: Comments on “medicolegal alcohol determination: Implications and consequences of irregularities in blood alcohol concentration vs. time curves.” J Analyt Toxic01 17:317, I993 (letter) 18. Watson PE: Total body water and blood alcohol levels: Updating

FRlEL ET AL.

the fundamentals, in Crow KE, Batt RD (eds): Human Metabolism of Alcohol, vol 1. Boca Raton, FL, CRC Press, 1989, pp 41-59 19. Friel PN, Logan B K Baer J: An evaluation of the reliability of Widmark calculations based on breath alcohol measurements. J Forensic Sci 40:91-94, 1995 20. Smith JD, Shaw U,Maini PK, Ward RJ, Peters TJ, Murray JD: Mathematical modeling of ethanol metabolism in normal subjects and chronic alcohol misusers. Alcohol Alcohol 28:25-32, 1993