Reproducibility of individual rates of ethanol metabolism in fasting subjects To estimate variability in ethanol metabolism, eight normal men received oral doses of 95% ethanol (1 nil/kg) after an overnight fast on each of 4 successive weeks. For each subject, slopes of linear decay curves of blood ethanol were highly reproduciblecoefficients of variation ranged from 3% for the least variable subject to 12% for the most variable subject (mean, 8%). Compared to this low intraindividual variation, interindividual variation was slightly higherthe mean coefficient of interindividual variation was 14%, with a range from 10% to 17%. A one-way ANOVA with repeat measures showed that on any one of four separate occasions of ethanol administration, the eight subjects differed from one another (p < 0.01), but that each subject remained similar from one test to another. The consumption of food before ethanol administration increased variability. Fever associated with upper respiratory infection also increased variability, but exercise did not. (CLIN PHARMACOL THER 1990;47:389-96.)
G. Thomas Passananti, PhD, Carol A. Wolff, MS, and Elliot S. Vesell, MD Hershey, Pa. Controversy persists concerning ethanol kinetics and
metabolism, despite intensive investigation."' Disagreements involve the questions of whether ethanol elimination conforms best to zero-order or to Michaelis-
Menten kinetics, whether it is described most suitably by a one- or two-compartment model, whether metabolism is mediated entirely by alcohol dehydrogenase or whether on occasion a highly inducible, but accessory, microsomal ethanol oxidizing system participates significantly and, finally, whether the ethanol elimination rate of a normal subject is a stable, highly reproducible trait or whether it fluctuates extensively and unpredictably.
Divergent results occurred in the past not only because ethanol elimination is inherently complex and dose-dependent, rendering it subject to differences in rates of gastrointestinal absorption of ethanol, but also because ethanol metabolism in the gastrointestinal tract and liver can be influenced by numerous host factors. These factors include age, sex, genetic constitution, diet, ingestion of other drugs, oral contraceptives, cigarette smoking, length of time ethanol has
been ingested, and the dose and dosage form of the
From the Department of Pharmacology, Pennsylvania State University College of Medicine. Received for publication Aug. 21, 1989; accepted Oct. 30, 1989. Reprint requests: Elliot S. Vesell, MD, Department of Pharmacology, Pennsylvania State University College of Medicine, PO Box 850, Hershey, PA 17033. 13/1/17848
ethanol administered.' Previous studies of subjects under nonuniform conditions demonstrated irreproducible results" that were attributed to the inherent nature of ethanol metabolism rather than to the environmental heterogeneity introduced both by the selection process and experimental protocols used.
A correct estimate of the relative magnitudes of intraindividual and interindividual variations in ethanol metabolism is significant for several reasons, including better understanding of mechanisms that reg-
ulate widely divergent human responses to alcohol. Despite its importance, this issue of the extent of variability in ethanol metabolism has not been satisfactorily settled, and it has not been recently addressed from the
perspective of using suitably selected subjects under appropriately uniform environmental conditions. For example, large intraindividual variations in ethanol metabolism were reported in several studies with nonfast-
ing subjects,' although food in the stomach at the time of ethanol ingestion increased variability in ethanol absorption9'° and metabolism.' Therefore, the present study readdressed this problem and used only environmentally homogeneous, carefully selected, agematched healthy men who fasted overnight before receiving their doses of ethanol. Our results differed from those of several previous reports" and revealed that under uniform controlled conditions, individual rates of ethanol metabolism are not unpredictable and do not fluctuate highly. Rather, individual rates are surprisingly reproducible. Variability between subjects was also small but was significantly larger than intraindividual variation.
389
390
Passananti,
Wolff,
CLIN PHARMACOL THER MARCH 1990
and Vesell
Table I. Reproducibility of the intercept (Cb in grams per 100 ml) of blood ethanol decay curve in normal subjects under stable conditions Subject
Age
W. E. J. P. K. K. E. G. R. M. M. G. A. L. R. H.
34 24 27 24 31
24 24 24
Weight (kg)
Week I
Week 2
86.0 66.5 80.5 75.0 71.0 83.5 82.5 93.0
0.08 0.09 0.10 0.10 0.10 0.10 0.09 0.12
0.10 0.09 0.09 0.09 0.10 0.09 0.09 0.12
Mean -± SD Coefficient of variation
0.0975 ± 0.012
0.0963 -± 0.011
11.95
11.02
Table II. Reproducibility of the slope (ko in grams per 100 ml/hr) of the blood ethanol decay curve in normal subjects under stable conditions Subject
W. E. J. P. K. K. E. G. R. M. M. G. A. L. R. H.
Age 34 24 27 24 31
24 24 24
Weight (kg)
86.0 66.5 80.5 75.0 71.0 83.5 82.5 93.0
Mean SD Coefficient of variation
MATERIAL AND METHODS Selection of subjects. Eight healthy men whose ages ranged from 24 to 34 years and whose weights ranged from 67 to 93 kg volunteered for this study. The objectives and risks of the study were explained to each subject (Table I). These subjects, who were students in our medical school, provided signed informed consent and were screened carefully. None had a history of serious illness, of regular consumption of any medication or alcoholic beverage, of smoking tobacco, or of chronic exposure to chemicals known to induce or inhibit hepatic drug-metabolizing enzymes. In addition, they appeared to be normal as shown by complete physical examination and routine blood chemistries, hemogram, and urinalysis. For 2 weeks before and during the study, subjects abstained from all medications and alcoholic beverages. Subjects were required to abstain from solid food for 12 hours before and 4 hours after receiving ethanol. Sample collection. At 9 AM on four successive Sat-
Week 1
Week 2
0.010 0.009 0.013 0.013 0.014 0.013
0.013 0.011 0.011 0.010 0.014 0.012 0.011 0.015
0.011 0.015
0.01225 ± 0.002
0.01213 ±- 0.002
16.76
14.24
urdays in January 1989 each subject drank, over a 15minute period, the same dose of ethanol: 1 ml/kg 95% ethanol in 250 ml ice-cold water. Subjects remained quiet and seated in the same room for the duration of sample collection. Beginning at 11 AM (2 hours after ethanol ingestion), seven blood samples of 10 ml each were collected every 1/2 hour by venipuncture of the antecubital vein into a sterile vacutainer blood collection tube that contained 20 mg potassium oxalate and 25 mg sodium fluoride.
Role of exercise. In four of our eight subjects, an additional measurement of ethanol metabolism was made several months after they completed the study described above to assess the effect of continuous moderate exercise for 2 hours on their rate of ethanol me-
tabolism. These four subjects (A. L., R. M., J. P., and W. E.) were accustomed to vigorous exercise on a daily basis. They walked on a motor driven treadmill for a period of 2 hours at a workload requiring 50% of their maximum aerobic capacity. This capacity was esti-
VOLUME 47 NUMBER 3
Reproducibility of ethanol metabolism
391
Coefficient Week 3
Week 4
Mean ± SD
of variation
0.10 0.09 0.10 0.09 0.09 0.10 0.10
0.10 0.10 0.09 0.09 0.09 0.09 0.08
10.53
0.11
0.11
0.0950 ± 0.010 0.0925 ± 0.005 0.0950 ± 0.006 0.0925 ± 0.005 0.0950 ± 0.006 0.0950 ± 0.006 0.0900 ± 0.008 0.1150 ± 0.006
0.0975 ± 0.007
0.0938 ± 0.009
7.25
5.40 6.08 5.40 6.08 6.08 9.07 5.02
9.77
Coefficient Week 3
Week 4
Mean ± SD
of variation
0.013
0.013 0.012 0.013 0.012 0.012 0.012 0.010 0.010
0.01225 ± 0.002 0.01075 ± 0.001 0.01225 ± 0.001 0.01150 ± 0.001 0.01325 ± 0.001 0.01300 ± 0.001 0.01125 ± 0.001 0.01475 ± 0.001
12.24 11.70
0.011
0.012 0.011 0.013 0.015 0.013 0.015
0.01288 ± 0.002
0.01225 ± 0.001
12.06
9.51
7.82 11.23
7.22 10.88 11.18 3.39
mated by the following equation to predict maximum heart rate and heart rate reserve:
as described previously. Rectal temperatures were taken before exercise and after 1 and 2 hours of exercise.
220 - Age - Resting heart rate (50%) +
ethanol four times at weekly intervals in a resting and fasting state, a final study was performed to assess the influence of food in the stomach on the individual rate of ethanol metabolism of a given subject. This study, performed 6 months later, was identical in design and details to the first study described earlier, except that subjects were not in the fasting state when ethanol was administered-each subject ate a moderate breakfast 1 hour before receiving his 9 AM dose of ethanol. The breakfast, which was eaten in the hospital cafeteria between 7:50 AM and 8:05 AM, was the same for all subjects: one large glass (700 ml) of orange juice, two fried eggs, three strips of bacon, two pieces of white toast with margarine, two cups of black coffee, and two glasses of ice water. Analysis of samples. Ethanol was assayed by head-
Resting heart rate = Predicted exercising heart rate
Each subject rode a stationary bike for 5 minutes at 150
kg m min' before the start of the exercise session. Exercise heart rates, monitored by palpation of the pulse, were checked every 30 minutes just before blood was drawn for measurement of ethanol concentration.
If necessary, the workload was adjusted to keep the subject working at 50% of his maximal heart rate. Rating of perceived exertion'' was also monitored and recorded in an effort to keep each subject exercising at the desired intensity. Ethanol (1 ml /kg 95% ethanol in 250 ml ice-cold water) was administered to these four fasting subjects at 9 AM. Exercise was started 2 hours later at 11 AM. While subjects exercised, blood specimens were drawn at the same seven 1/2-hour intervals
Role of food. In the eight subjects who received
CLIN PHARMACOL THER MARCH 1990
392 Passananti, Wolff, and Vesell .10
.10
A.L.
.08
.08
.06
.06
.04
.04
.02
.02
0
o
.10
.10
E.G.
.08
.08
.06
.06
.04
.04
.02
.02
o
o
.10
.10
..08
.08
.06
^
.04
.02
.02
o
o
.10 -
0
J.P.
R.M.
.06
.04
0
K.K.
.10 -
M.G.
1
t
i
1
2
3
o
I
4
5
0
W.E.
1
2
3
4
5
HOURS
Fig. 1. Four blood ethanol decay curves for each of eight fasting subjects who received an oral dose of ethanol on four separate occasions. R. H. received ethanol on a fifth occasion (indicated by open circles) because he had an upper respiratory infection when the curve with the shallow slope was obtained.
space gas chromatography according to the method of Goldbaum.' One milliliter of blood was added to 4 ml aqueous solution of internal standard (16.6 mmol/L npropanol) in a screw-capped vial with a puncture-type silicone septum. After a 30-minute incubation period at 60° C, 2 ml of the vapor phase was removed with
an air-tight disposable syringe, 1 ml was expelled, and
1 ml was injected into a Hewlett-Packard 5830A gas chromatograph with a 18850A GC terminal and flame ionization detector (Hewlett-Packard Company, Palo Alto, Calif.). The operating conditions for the ethanol determination were as follows: column tem-
VOLUME 47 NUMBER 3
perature, 90° C; flash temperature, 100° C; and detector temperature, 250° C. A 6-foot coiled chromatographic column, packed with GP carbopack C 60 / 80 and 0.2%
carbowax 1500, was used. This material is commercially available from Supelco Inc., Bellefonte, Pa.
Standards were prepared by addition of known amounts of ethanol to whole human blood that contained sodium fluoride-potassium oxalate in the same concentration that was present in the samples:8 Calibration curves were linear up to 0.2% ethanol concentration.
RESULTS The main observations in this study on intraindividual versus interindividual variability in ethanol metabolism of fasting subjects appear in Tables I and II and Fig. 1. Table I shows the values for Cb, the ethanol concentration in grams per 100 ml blood at time zero of ethanol ingestion. The Cb was estimated by extrap-
olation to the y intercept of the linear portion of the ethanol decay curve in blood (Fig. 1). Table II shows the values for k., the slope of the ethanol decay curve expressed in grams per 100 ml blood per hour. Both Cb and ko were highly reproducible in each subject on the four occasions on which they were measured, as indicated by the low coefficient of variation, which had
a mean value of 8% for ko, with a range from 3% to 12%, and a mean value of 7% for Cb, with a range from 5% to 11%. Also, Fig. 1 shows close similarity between the four ethanol decay curves obtained for each subject. Compared to the high reproducibility of ethanol decay curves in each normal subject, ethanol metabolism
varied more extensively between subjects. The extent of this interindividual variation was 75% more than the magnitude of intraindividual variation, as indicated by a mean value of 14% for the coefficient of interindividual variation of ko, with a range of 10% to 17% on the four separate occasions it was measured. The mean coefficient of interindividual variation for Cb was 10%, with a range from 7% to 12%. A one-way ANOVA with repeat measures for both ko and Cb showed that on any one of the four separate occasions of ethanol administration the eight subjects differed from one another (p < 0.01), but each subject remained similar from one test to another. Although Fig. 1 shows high reproducibility of ethanol decay curves in a given subject on each of the four occasions on which ethanol was measured, one notable exception occurred. In subject R. H. the slope of one
curve was clearly less steep than on the other three
Reproducibility of ethanol metabolism
393
occasions. That curve was obtained on the fourth and final occasion of ethanol administration. The subject arrived for the test with an upper respiratory infection, rhinorrhea, and fever. Rather than send him away without completing the study, we administered the ethanol. To determine whether the shallow slope of the curve obtained on this occasion (0.010 compared with a value of 0.015 obtained for each of the other three trials) was indeed an exception, we administered ethanol to R. H. on a fifth occasion 2 months after the fourth measurement. The 0.014 slope of the repeat, shown in Fig. 1, suggests that fourth measurement for R. H. was indeed unusual (and probably related to his upper respiratory infection) compared with his highly reproducible values
for ethanol metabolism obtained under stable, unperturbed environmental conditions. Thus for purposes of calculation of variability, the fifth measurement of Cb and ko for R. H. was used rather than his aberrant fourth measurement, which appeared to be associated with his upper respiratory infection. Effects on ethanol kinetics of food in the stomach at
the time of ethanol ingestion are shown in Fig. 2. In four subjects the ethanol concentrations were higher, whereas in the other four subjects they were lower than
when ethanol was ingested in a fasting state (Fig. 2). Three indications that food in the stomach at the time of ethanol ingestion increased variability in ethanol kinetics for subjects K. K., E. G., J. P., and W. E. are the altered slopes of ethanol decay curves of these subjects, several ethanol concentrations more than 2 standard deviations away from mean fasting values, and several 2-hour ethanol concentrations below the level expected from the decay curve slope, suggesting incomplete gastrointestinal absorption of ethanol at this initial time. After continuous exercise on a treadmill for 2 hours, beginning 2 hours after ethanol ingestion, the ethanol decay curves obtained for the four subjects (A. L., R. M., J. P., and W. E.) were unchanged from those shown in Fig. 1 under resting conditions. The rectal
temperatures of these four subjects rose 1.0° C to 1.5° C during exercise from the normal values measured before exercise.
DISCUSSION High reproducibility of individual rates of ethanol metabolism appears to contradict earlier reports" that claimed marked intraindividual variability. This apparent discrepancy can be explained by critical differences between our study and earlier ones in design and condition of subjects. Specifically, we used fasting sub-
CLIN PHARMACOL THER MARCH 1990
394 Passananti, Wolff, and Vesell
10
AL.
08
.10
K.K.
08 o o
06
06 o
o
04
04 °
02
02 o
10
E.G.
06
06
o o
04
oO
r
.02
C.5
o
E o
.10 -
et
3C co .c
.08
o
o
o
o
.r.
e
J.P.
08
08
i
10-
o
-
a
04
02
o
-
R.N.
i
i
R.M.
.08 1
°
.06
.10
o
i i
.06
io
WI
4 o
; o
I o
.04
04
02
02
o
o
o
10
M.G.
.08
.10
08 o
o o
.06
06'
o
I
r 7
; o
ro
04
o
.04
.02
o
W.E.
I o
E
.02
;
;
Hours
4
5
2
2
4
5
Hours
Fig. 2. Mean blood ethanol concentrations of four separate studies appear as solid circles; 1 SD is indicated by vertical line. Open circles indicate ethanol concentrations in the same subjects on another study in which they received breakfast 1 hour before ethanol administration. Note greater variability after food, indicated in some subjects by different slopes of decay curves, several ethanol concentrations more than 2 SD from the mean values, and incomplete ethanol absorption at several 2 hour points.
VOLUME 47 NUMBER 3
jects, whereas subjects in earlier studies were administered ethanol shortly after a meal.' Food in the stomach at the time of ethanol administration retards ethanol absorption, thereby affecting ethanol elimination,
which is dose-dependent.' One part of our study (Fig. 2) examined effects of
Reproducibility of ethanol metabolism
395
elevated temperature by itself is insufficient to alter ethanol kinetics. The central body temperature of four of our subjects was raised by 1.0° C to 1.5° C during moderate physical exercise (recognized as a normal occurrence during such exertion20-22), but no change oc-
curred in their ethanol kinetics measured at rest with
food consumption shortly before ethanol administration on variability in ethanol kinetics. Food in the stomach altered ethanol elimination rates and concentrations in half of our subjects, compared with their fasting values
normal body temperature. Thus some aspect of the viral infection other than fever (possibly an alteration in hepatic function) may affect ethanol metabolism. In this regard, interferon elevations during the course of certain
(Fig. 2, subjects K. K., E. G., J. P., and W. E.). High reproducibility of individual rates of ethanol
viral infections inhibit hepatic drug metabolism.' Third, claims that several hundred twins must be studied to reach statistically meaningful conclusions on the role of genetic factors in the regulation of interindividual differences in rates of ethanol metabolism' are based on twins that exhibited large intraindividual variations in ethanol metabolism. As the present study suggests, the magnitude of intraindividual variation can depend on the environmental heterogeneity of the subjects selected and the experimental design used. Although large numbers of twins were required in
metabolism under the conditions of this study has several implications. First, because variability of ethanol kinetics among subjects under uniform environmental conditions exceeded that within subjects (Tables I and II), genetic factors merit consideration as contributing elements. Genetic control of interindividual variations in ethanol metabolism was previously suggested by the results of several twin studies that revealed less variability in ethanol metabolism within monozygotic than within dizygotic twins.6'9 The role of genetic factors in the regulation of interindividual variations in ethanol
metabolism was confirmed in subsequent studies by Kopun and Propping' and by Martin et al.3 However, these latter studies ascribed a much smaller proportion of the total interindividual variability to genetic factors and a much larger portion to environmental factors. This was probably because of the inclusion of environmentally perturbed subjects in these studies. Specifically, subjects consumed ethanol regularly and ethanol was also given shortly after a meal, two host factors recognized to increase variability in ethanol kinetics. Thus, although the results of the present study indicated that rates of ethanol metabolism in individual subjects are highly reproducible under uniform, care-
the study of Martin et al.3 because of the environmental heterogeneity of subjects in that study with consequent large intraindividual variation, an advantage in the use
of more environmentally homogeneous subjects and conditions is that fewer subjects are required. Fewer subjects are needed because their rates of ethanol metabolism are highly reproducible (a mean coefficient of variation of 8%, with a range from 3% to 12% for k0 in our subjects). Therefore, extra effort spent in selecting environmentally homogeneous subjects and in imposing uniform environmental conditions during the study is compensated for because fewer subjects are required to reach statistical significance. Finally, in studies designed to measure variability in drug metabolism among and between subjects, claims
fully controlled environmental conditions and although twin studies conducted under such conditions have sug-
of large intraindividual variations should be assessed in
gested that genetic factors are largely responsible for differences among subjects in these rates," the rate of ethanol metabolism in a given subject is exceedingly sensitive to environmental perturbation. Second, in addition to the host factors of long-term ethanol ingestion and food in the stomach, a novel environmental factor that affects ethanol kinetics was suggested by a chance occurrence during the course of our study. When subject R. H. had an upper respiratory
thereby assuring that all subjects who receive a drug do so under appropriately uniform environmental condi-
light of carefully applied rigorous selection criteria,
tions. Otherwise, lack of reproducible kinetics in a subject should be assumed to reflect the particular conditions under which that subject was investigated, rather than some inherent property of hepatic drug metabolism. References
infection, the slope of his ethanol decay curve de-
Wagner JG, Patel JA. Variations in absorption and elim-
creased to 0.010 from values of 0.015 that were measured on three previous occasions and a value of 0.014 measured subsequently (Fig. 1 and Table I). However,
ination rates of ethyl alcohol in a single subject. Res Commun Chem Pathol Pharmacol 1972;4:61-76. Wagner JG. Intrasubject variation in elimination half-
CL1N PHARMACOL THER MARCH 1990
396 Passananti, Wolff, and Vesell lives of drugs which are appreciably metabolized. J Pharmacokinet Biopharm 1973;1:165-73.
Weidler DJ, Wagner JO. Blood ethanol concentrations during and following constant-rate intravenous infusion
Martin NG, Perl J, Oakeshott JG, Gibson JB, Starmer GA, Wilks AV. A twin study of ethanol metabolism.
of alcohol. CLIN PHARMACOL THER 1976;19:213-23.
Behav Genet 1985;15:93-109. Martin NG, Eaves LJ, Kearsey MJ, Davies P. The power of the classical twin study. Heredity (Lond) 1978;40:97-
O'Neill B, Williams AF, Dubowski KM. Variability in blood alcohol concentrations. Implications for estimating individual results. J Stud Alcohol 1983;44:222-30. Sedman AJ, Wilkinson PK, Wagner JO. Concentrations
116.
of ethanol in two segments of the vascular system.
Martin NG, Oakeshott JO, Gibson JB, Starmer GA, Perl J, Wilks AV. A twin study of psychomotor and physio-
J Forensic Sci 1975;21:315-22. Lieber CS. Biochemical and molecular basis of alcoholinduced injury to liver and other tissues. N Engl J Med 1988;319:1639-50. Borg GV. Psychophysical basis of perceived exertion. Med Sci Sports Exercise 1982;14:377-81. Goldbaum LR. Gas chromatographic method for the measurement of ethyl alcohol. In: Sunderman FW, Sunderman FW Jr, eds. Laboratory diagnosis of diseases caused
logical responses to an acute dose of alcohol. Behav Genet 1985;15:305-47. Luth KF. Untersuchungen uber die Alkoholblutconzen-
tration nach Alkoholgaben bei 10 eineiigen und 10 zweieigen Zwillingspaaren. Dtsch Z Gerichtl Med 1939;32:145-64. Kopun M, Propping P. The kinetics of ethanol absorption and elimination in twins supplemented by repetitive ex-
by toxic agents. St Louis: Warren H Green, 1970;
periments in single subjects. Eur J Clin Pharmacol
312-5.
1977;11:337-44. Holford NHG. Clinical pharmacokinetics of ethanol. Clin Pharmacolcinet 1987;13:273-92.
Vesell ES, Page JO, Passananti GT. Genetic and environmental factors affecting ethanol metabolism in man.
Lin YJ, Weidler DJ, Garg DC, Wagner JO. Effects of solid food on blood levels of alcohol in man. Res Com-
Sawka MN, Knowlton RO, Critz JB. Thermal and circulatory responses to repeated bouts of prolonged running. Med Sci Sports 1979;11:177-80. Sawka MN, Pimental NA, Pandolf KB. Thermoregula-
mun Chem Pathol Pharmacol 1976;13:713-22. Sedman AJ, Wilkinson PK, Sakmar E, Weidler Di, Wag-
ner JG. Food effects on absorption and metabolism of alcohol. J Stud Alcohol 1976;37:1197-214. Wilkinson PK, Sedman AJ, Sakmar E, Kay DR, Wagner JO. Pharmacokinetics of ethanol after oral administra-
tion in the fasting state. J Pharmacokinet Biopharm 1977;5:207-24.
Rangno RE, Kreeft JH, Sitar DS. Ethanol dosedependent elimination: Michaelis-Menten classical kinetic analysis. Br J Clin Pharmacol 1981;12:667-73.
Wilkinson PK, Sedman AJ, Sakmar E, Earhart RH,
CLIN PHARMACOL THER 1971;12:192-201.
tory responses to upper body exercise. EurJ Appl Physiol 1984;52:230-4. Sawka MN, Young AJ, Francesconi RP, Muza SR, Pandolf KB. Thermoregulatory and blood responses during exercise at graded hypohydration levels. J Appl Physiol 1985;59:1394-401. Mannering GJ, Deloria LB. The pharmacology and toxicology of the interferons: an overview. Annu Rev Pharmacol Toxicol 1986;26:480-7.
G. Thomas Passananti, PhD, Carol A. Wolff, MS, and Elliot S. Vesell, MD Hershey, Pa. Controversy persists concerning ethanol kinetics and
metabolism, despite intensive investigation."' Disagreements involve the questions of whether ethanol elimination conforms best to zero-order or to Michaelis-
Menten kinetics, whether it is described most suitably by a one- or two-compartment model, whether metabolism is mediated entirely by alcohol dehydrogenase or whether on occasion a highly inducible, but accessory, microsomal ethanol oxidizing system participates significantly and, finally, whether the ethanol elimination rate of a normal subject is a stable, highly reproducible trait or whether it fluctuates extensively and unpredictably.
Divergent results occurred in the past not only because ethanol elimination is inherently complex and dose-dependent, rendering it subject to differences in rates of gastrointestinal absorption of ethanol, but also because ethanol metabolism in the gastrointestinal tract and liver can be influenced by numerous host factors. These factors include age, sex, genetic constitution, diet, ingestion of other drugs, oral contraceptives, cigarette smoking, length of time ethanol has
been ingested, and the dose and dosage form of the
From the Department of Pharmacology, Pennsylvania State University College of Medicine. Received for publication Aug. 21, 1989; accepted Oct. 30, 1989. Reprint requests: Elliot S. Vesell, MD, Department of Pharmacology, Pennsylvania State University College of Medicine, PO Box 850, Hershey, PA 17033. 13/1/17848
ethanol administered.' Previous studies of subjects under nonuniform conditions demonstrated irreproducible results" that were attributed to the inherent nature of ethanol metabolism rather than to the environmental heterogeneity introduced both by the selection process and experimental protocols used.
A correct estimate of the relative magnitudes of intraindividual and interindividual variations in ethanol metabolism is significant for several reasons, including better understanding of mechanisms that reg-
ulate widely divergent human responses to alcohol. Despite its importance, this issue of the extent of variability in ethanol metabolism has not been satisfactorily settled, and it has not been recently addressed from the
perspective of using suitably selected subjects under appropriately uniform environmental conditions. For example, large intraindividual variations in ethanol metabolism were reported in several studies with nonfast-
ing subjects,' although food in the stomach at the time of ethanol ingestion increased variability in ethanol absorption9'° and metabolism.' Therefore, the present study readdressed this problem and used only environmentally homogeneous, carefully selected, agematched healthy men who fasted overnight before receiving their doses of ethanol. Our results differed from those of several previous reports" and revealed that under uniform controlled conditions, individual rates of ethanol metabolism are not unpredictable and do not fluctuate highly. Rather, individual rates are surprisingly reproducible. Variability between subjects was also small but was significantly larger than intraindividual variation.
389
390
Passananti,
Wolff,
CLIN PHARMACOL THER MARCH 1990
and Vesell
Table I. Reproducibility of the intercept (Cb in grams per 100 ml) of blood ethanol decay curve in normal subjects under stable conditions Subject
Age
W. E. J. P. K. K. E. G. R. M. M. G. A. L. R. H.
34 24 27 24 31
24 24 24
Weight (kg)
Week I
Week 2
86.0 66.5 80.5 75.0 71.0 83.5 82.5 93.0
0.08 0.09 0.10 0.10 0.10 0.10 0.09 0.12
0.10 0.09 0.09 0.09 0.10 0.09 0.09 0.12
Mean -± SD Coefficient of variation
0.0975 ± 0.012
0.0963 -± 0.011
11.95
11.02
Table II. Reproducibility of the slope (ko in grams per 100 ml/hr) of the blood ethanol decay curve in normal subjects under stable conditions Subject
W. E. J. P. K. K. E. G. R. M. M. G. A. L. R. H.
Age 34 24 27 24 31
24 24 24
Weight (kg)
86.0 66.5 80.5 75.0 71.0 83.5 82.5 93.0
Mean SD Coefficient of variation
MATERIAL AND METHODS Selection of subjects. Eight healthy men whose ages ranged from 24 to 34 years and whose weights ranged from 67 to 93 kg volunteered for this study. The objectives and risks of the study were explained to each subject (Table I). These subjects, who were students in our medical school, provided signed informed consent and were screened carefully. None had a history of serious illness, of regular consumption of any medication or alcoholic beverage, of smoking tobacco, or of chronic exposure to chemicals known to induce or inhibit hepatic drug-metabolizing enzymes. In addition, they appeared to be normal as shown by complete physical examination and routine blood chemistries, hemogram, and urinalysis. For 2 weeks before and during the study, subjects abstained from all medications and alcoholic beverages. Subjects were required to abstain from solid food for 12 hours before and 4 hours after receiving ethanol. Sample collection. At 9 AM on four successive Sat-
Week 1
Week 2
0.010 0.009 0.013 0.013 0.014 0.013
0.013 0.011 0.011 0.010 0.014 0.012 0.011 0.015
0.011 0.015
0.01225 ± 0.002
0.01213 ±- 0.002
16.76
14.24
urdays in January 1989 each subject drank, over a 15minute period, the same dose of ethanol: 1 ml/kg 95% ethanol in 250 ml ice-cold water. Subjects remained quiet and seated in the same room for the duration of sample collection. Beginning at 11 AM (2 hours after ethanol ingestion), seven blood samples of 10 ml each were collected every 1/2 hour by venipuncture of the antecubital vein into a sterile vacutainer blood collection tube that contained 20 mg potassium oxalate and 25 mg sodium fluoride.
Role of exercise. In four of our eight subjects, an additional measurement of ethanol metabolism was made several months after they completed the study described above to assess the effect of continuous moderate exercise for 2 hours on their rate of ethanol me-
tabolism. These four subjects (A. L., R. M., J. P., and W. E.) were accustomed to vigorous exercise on a daily basis. They walked on a motor driven treadmill for a period of 2 hours at a workload requiring 50% of their maximum aerobic capacity. This capacity was esti-
VOLUME 47 NUMBER 3
Reproducibility of ethanol metabolism
391
Coefficient Week 3
Week 4
Mean ± SD
of variation
0.10 0.09 0.10 0.09 0.09 0.10 0.10
0.10 0.10 0.09 0.09 0.09 0.09 0.08
10.53
0.11
0.11
0.0950 ± 0.010 0.0925 ± 0.005 0.0950 ± 0.006 0.0925 ± 0.005 0.0950 ± 0.006 0.0950 ± 0.006 0.0900 ± 0.008 0.1150 ± 0.006
0.0975 ± 0.007
0.0938 ± 0.009
7.25
5.40 6.08 5.40 6.08 6.08 9.07 5.02
9.77
Coefficient Week 3
Week 4
Mean ± SD
of variation
0.013
0.013 0.012 0.013 0.012 0.012 0.012 0.010 0.010
0.01225 ± 0.002 0.01075 ± 0.001 0.01225 ± 0.001 0.01150 ± 0.001 0.01325 ± 0.001 0.01300 ± 0.001 0.01125 ± 0.001 0.01475 ± 0.001
12.24 11.70
0.011
0.012 0.011 0.013 0.015 0.013 0.015
0.01288 ± 0.002
0.01225 ± 0.001
12.06
9.51
7.82 11.23
7.22 10.88 11.18 3.39
mated by the following equation to predict maximum heart rate and heart rate reserve:
as described previously. Rectal temperatures were taken before exercise and after 1 and 2 hours of exercise.
220 - Age - Resting heart rate (50%) +
ethanol four times at weekly intervals in a resting and fasting state, a final study was performed to assess the influence of food in the stomach on the individual rate of ethanol metabolism of a given subject. This study, performed 6 months later, was identical in design and details to the first study described earlier, except that subjects were not in the fasting state when ethanol was administered-each subject ate a moderate breakfast 1 hour before receiving his 9 AM dose of ethanol. The breakfast, which was eaten in the hospital cafeteria between 7:50 AM and 8:05 AM, was the same for all subjects: one large glass (700 ml) of orange juice, two fried eggs, three strips of bacon, two pieces of white toast with margarine, two cups of black coffee, and two glasses of ice water. Analysis of samples. Ethanol was assayed by head-
Resting heart rate = Predicted exercising heart rate
Each subject rode a stationary bike for 5 minutes at 150
kg m min' before the start of the exercise session. Exercise heart rates, monitored by palpation of the pulse, were checked every 30 minutes just before blood was drawn for measurement of ethanol concentration.
If necessary, the workload was adjusted to keep the subject working at 50% of his maximal heart rate. Rating of perceived exertion'' was also monitored and recorded in an effort to keep each subject exercising at the desired intensity. Ethanol (1 ml /kg 95% ethanol in 250 ml ice-cold water) was administered to these four fasting subjects at 9 AM. Exercise was started 2 hours later at 11 AM. While subjects exercised, blood specimens were drawn at the same seven 1/2-hour intervals
Role of food. In the eight subjects who received
CLIN PHARMACOL THER MARCH 1990
392 Passananti, Wolff, and Vesell .10
.10
A.L.
.08
.08
.06
.06
.04
.04
.02
.02
0
o
.10
.10
E.G.
.08
.08
.06
.06
.04
.04
.02
.02
o
o
.10
.10
..08
.08
.06
^
.04
.02
.02
o
o
.10 -
0
J.P.
R.M.
.06
.04
0
K.K.
.10 -
M.G.
1
t
i
1
2
3
o
I
4
5
0
W.E.
1
2
3
4
5
HOURS
Fig. 1. Four blood ethanol decay curves for each of eight fasting subjects who received an oral dose of ethanol on four separate occasions. R. H. received ethanol on a fifth occasion (indicated by open circles) because he had an upper respiratory infection when the curve with the shallow slope was obtained.
space gas chromatography according to the method of Goldbaum.' One milliliter of blood was added to 4 ml aqueous solution of internal standard (16.6 mmol/L npropanol) in a screw-capped vial with a puncture-type silicone septum. After a 30-minute incubation period at 60° C, 2 ml of the vapor phase was removed with
an air-tight disposable syringe, 1 ml was expelled, and
1 ml was injected into a Hewlett-Packard 5830A gas chromatograph with a 18850A GC terminal and flame ionization detector (Hewlett-Packard Company, Palo Alto, Calif.). The operating conditions for the ethanol determination were as follows: column tem-
VOLUME 47 NUMBER 3
perature, 90° C; flash temperature, 100° C; and detector temperature, 250° C. A 6-foot coiled chromatographic column, packed with GP carbopack C 60 / 80 and 0.2%
carbowax 1500, was used. This material is commercially available from Supelco Inc., Bellefonte, Pa.
Standards were prepared by addition of known amounts of ethanol to whole human blood that contained sodium fluoride-potassium oxalate in the same concentration that was present in the samples:8 Calibration curves were linear up to 0.2% ethanol concentration.
RESULTS The main observations in this study on intraindividual versus interindividual variability in ethanol metabolism of fasting subjects appear in Tables I and II and Fig. 1. Table I shows the values for Cb, the ethanol concentration in grams per 100 ml blood at time zero of ethanol ingestion. The Cb was estimated by extrap-
olation to the y intercept of the linear portion of the ethanol decay curve in blood (Fig. 1). Table II shows the values for k., the slope of the ethanol decay curve expressed in grams per 100 ml blood per hour. Both Cb and ko were highly reproducible in each subject on the four occasions on which they were measured, as indicated by the low coefficient of variation, which had
a mean value of 8% for ko, with a range from 3% to 12%, and a mean value of 7% for Cb, with a range from 5% to 11%. Also, Fig. 1 shows close similarity between the four ethanol decay curves obtained for each subject. Compared to the high reproducibility of ethanol decay curves in each normal subject, ethanol metabolism
varied more extensively between subjects. The extent of this interindividual variation was 75% more than the magnitude of intraindividual variation, as indicated by a mean value of 14% for the coefficient of interindividual variation of ko, with a range of 10% to 17% on the four separate occasions it was measured. The mean coefficient of interindividual variation for Cb was 10%, with a range from 7% to 12%. A one-way ANOVA with repeat measures for both ko and Cb showed that on any one of the four separate occasions of ethanol administration the eight subjects differed from one another (p < 0.01), but each subject remained similar from one test to another. Although Fig. 1 shows high reproducibility of ethanol decay curves in a given subject on each of the four occasions on which ethanol was measured, one notable exception occurred. In subject R. H. the slope of one
curve was clearly less steep than on the other three
Reproducibility of ethanol metabolism
393
occasions. That curve was obtained on the fourth and final occasion of ethanol administration. The subject arrived for the test with an upper respiratory infection, rhinorrhea, and fever. Rather than send him away without completing the study, we administered the ethanol. To determine whether the shallow slope of the curve obtained on this occasion (0.010 compared with a value of 0.015 obtained for each of the other three trials) was indeed an exception, we administered ethanol to R. H. on a fifth occasion 2 months after the fourth measurement. The 0.014 slope of the repeat, shown in Fig. 1, suggests that fourth measurement for R. H. was indeed unusual (and probably related to his upper respiratory infection) compared with his highly reproducible values
for ethanol metabolism obtained under stable, unperturbed environmental conditions. Thus for purposes of calculation of variability, the fifth measurement of Cb and ko for R. H. was used rather than his aberrant fourth measurement, which appeared to be associated with his upper respiratory infection. Effects on ethanol kinetics of food in the stomach at
the time of ethanol ingestion are shown in Fig. 2. In four subjects the ethanol concentrations were higher, whereas in the other four subjects they were lower than
when ethanol was ingested in a fasting state (Fig. 2). Three indications that food in the stomach at the time of ethanol ingestion increased variability in ethanol kinetics for subjects K. K., E. G., J. P., and W. E. are the altered slopes of ethanol decay curves of these subjects, several ethanol concentrations more than 2 standard deviations away from mean fasting values, and several 2-hour ethanol concentrations below the level expected from the decay curve slope, suggesting incomplete gastrointestinal absorption of ethanol at this initial time. After continuous exercise on a treadmill for 2 hours, beginning 2 hours after ethanol ingestion, the ethanol decay curves obtained for the four subjects (A. L., R. M., J. P., and W. E.) were unchanged from those shown in Fig. 1 under resting conditions. The rectal
temperatures of these four subjects rose 1.0° C to 1.5° C during exercise from the normal values measured before exercise.
DISCUSSION High reproducibility of individual rates of ethanol metabolism appears to contradict earlier reports" that claimed marked intraindividual variability. This apparent discrepancy can be explained by critical differences between our study and earlier ones in design and condition of subjects. Specifically, we used fasting sub-
CLIN PHARMACOL THER MARCH 1990
394 Passananti, Wolff, and Vesell
10
AL.
08
.10
K.K.
08 o o
06
06 o
o
04
04 °
02
02 o
10
E.G.
06
06
o o
04
oO
r
.02
C.5
o
E o
.10 -
et
3C co .c
.08
o
o
o
o
.r.
e
J.P.
08
08
i
10-
o
-
a
04
02
o
-
R.N.
i
i
R.M.
.08 1
°
.06
.10
o
i i
.06
io
WI
4 o
; o
I o
.04
04
02
02
o
o
o
10
M.G.
.08
.10
08 o
o o
.06
06'
o
I
r 7
; o
ro
04
o
.04
.02
o
W.E.
I o
E
.02
;
;
Hours
4
5
2
2
4
5
Hours
Fig. 2. Mean blood ethanol concentrations of four separate studies appear as solid circles; 1 SD is indicated by vertical line. Open circles indicate ethanol concentrations in the same subjects on another study in which they received breakfast 1 hour before ethanol administration. Note greater variability after food, indicated in some subjects by different slopes of decay curves, several ethanol concentrations more than 2 SD from the mean values, and incomplete ethanol absorption at several 2 hour points.
VOLUME 47 NUMBER 3
jects, whereas subjects in earlier studies were administered ethanol shortly after a meal.' Food in the stomach at the time of ethanol administration retards ethanol absorption, thereby affecting ethanol elimination,
which is dose-dependent.' One part of our study (Fig. 2) examined effects of
Reproducibility of ethanol metabolism
395
elevated temperature by itself is insufficient to alter ethanol kinetics. The central body temperature of four of our subjects was raised by 1.0° C to 1.5° C during moderate physical exercise (recognized as a normal occurrence during such exertion20-22), but no change oc-
curred in their ethanol kinetics measured at rest with
food consumption shortly before ethanol administration on variability in ethanol kinetics. Food in the stomach altered ethanol elimination rates and concentrations in half of our subjects, compared with their fasting values
normal body temperature. Thus some aspect of the viral infection other than fever (possibly an alteration in hepatic function) may affect ethanol metabolism. In this regard, interferon elevations during the course of certain
(Fig. 2, subjects K. K., E. G., J. P., and W. E.). High reproducibility of individual rates of ethanol
viral infections inhibit hepatic drug metabolism.' Third, claims that several hundred twins must be studied to reach statistically meaningful conclusions on the role of genetic factors in the regulation of interindividual differences in rates of ethanol metabolism' are based on twins that exhibited large intraindividual variations in ethanol metabolism. As the present study suggests, the magnitude of intraindividual variation can depend on the environmental heterogeneity of the subjects selected and the experimental design used. Although large numbers of twins were required in
metabolism under the conditions of this study has several implications. First, because variability of ethanol kinetics among subjects under uniform environmental conditions exceeded that within subjects (Tables I and II), genetic factors merit consideration as contributing elements. Genetic control of interindividual variations in ethanol metabolism was previously suggested by the results of several twin studies that revealed less variability in ethanol metabolism within monozygotic than within dizygotic twins.6'9 The role of genetic factors in the regulation of interindividual variations in ethanol
metabolism was confirmed in subsequent studies by Kopun and Propping' and by Martin et al.3 However, these latter studies ascribed a much smaller proportion of the total interindividual variability to genetic factors and a much larger portion to environmental factors. This was probably because of the inclusion of environmentally perturbed subjects in these studies. Specifically, subjects consumed ethanol regularly and ethanol was also given shortly after a meal, two host factors recognized to increase variability in ethanol kinetics. Thus, although the results of the present study indicated that rates of ethanol metabolism in individual subjects are highly reproducible under uniform, care-
the study of Martin et al.3 because of the environmental heterogeneity of subjects in that study with consequent large intraindividual variation, an advantage in the use
of more environmentally homogeneous subjects and conditions is that fewer subjects are required. Fewer subjects are needed because their rates of ethanol metabolism are highly reproducible (a mean coefficient of variation of 8%, with a range from 3% to 12% for k0 in our subjects). Therefore, extra effort spent in selecting environmentally homogeneous subjects and in imposing uniform environmental conditions during the study is compensated for because fewer subjects are required to reach statistical significance. Finally, in studies designed to measure variability in drug metabolism among and between subjects, claims
fully controlled environmental conditions and although twin studies conducted under such conditions have sug-
of large intraindividual variations should be assessed in
gested that genetic factors are largely responsible for differences among subjects in these rates," the rate of ethanol metabolism in a given subject is exceedingly sensitive to environmental perturbation. Second, in addition to the host factors of long-term ethanol ingestion and food in the stomach, a novel environmental factor that affects ethanol kinetics was suggested by a chance occurrence during the course of our study. When subject R. H. had an upper respiratory
thereby assuring that all subjects who receive a drug do so under appropriately uniform environmental condi-
light of carefully applied rigorous selection criteria,
tions. Otherwise, lack of reproducible kinetics in a subject should be assumed to reflect the particular conditions under which that subject was investigated, rather than some inherent property of hepatic drug metabolism. References
infection, the slope of his ethanol decay curve de-
Wagner JG, Patel JA. Variations in absorption and elim-
creased to 0.010 from values of 0.015 that were measured on three previous occasions and a value of 0.014 measured subsequently (Fig. 1 and Table I). However,
ination rates of ethyl alcohol in a single subject. Res Commun Chem Pathol Pharmacol 1972;4:61-76. Wagner JG. Intrasubject variation in elimination half-
CL1N PHARMACOL THER MARCH 1990
396 Passananti, Wolff, and Vesell lives of drugs which are appreciably metabolized. J Pharmacokinet Biopharm 1973;1:165-73.
Weidler DJ, Wagner JO. Blood ethanol concentrations during and following constant-rate intravenous infusion
Martin NG, Perl J, Oakeshott JG, Gibson JB, Starmer GA, Wilks AV. A twin study of ethanol metabolism.
of alcohol. CLIN PHARMACOL THER 1976;19:213-23.
Behav Genet 1985;15:93-109. Martin NG, Eaves LJ, Kearsey MJ, Davies P. The power of the classical twin study. Heredity (Lond) 1978;40:97-
O'Neill B, Williams AF, Dubowski KM. Variability in blood alcohol concentrations. Implications for estimating individual results. J Stud Alcohol 1983;44:222-30. Sedman AJ, Wilkinson PK, Wagner JO. Concentrations
116.
of ethanol in two segments of the vascular system.
Martin NG, Oakeshott JO, Gibson JB, Starmer GA, Perl J, Wilks AV. A twin study of psychomotor and physio-
J Forensic Sci 1975;21:315-22. Lieber CS. Biochemical and molecular basis of alcoholinduced injury to liver and other tissues. N Engl J Med 1988;319:1639-50. Borg GV. Psychophysical basis of perceived exertion. Med Sci Sports Exercise 1982;14:377-81. Goldbaum LR. Gas chromatographic method for the measurement of ethyl alcohol. In: Sunderman FW, Sunderman FW Jr, eds. Laboratory diagnosis of diseases caused
logical responses to an acute dose of alcohol. Behav Genet 1985;15:305-47. Luth KF. Untersuchungen uber die Alkoholblutconzen-
tration nach Alkoholgaben bei 10 eineiigen und 10 zweieigen Zwillingspaaren. Dtsch Z Gerichtl Med 1939;32:145-64. Kopun M, Propping P. The kinetics of ethanol absorption and elimination in twins supplemented by repetitive ex-
by toxic agents. St Louis: Warren H Green, 1970;
periments in single subjects. Eur J Clin Pharmacol
312-5.
1977;11:337-44. Holford NHG. Clinical pharmacokinetics of ethanol. Clin Pharmacolcinet 1987;13:273-92.
Vesell ES, Page JO, Passananti GT. Genetic and environmental factors affecting ethanol metabolism in man.
Lin YJ, Weidler DJ, Garg DC, Wagner JO. Effects of solid food on blood levels of alcohol in man. Res Com-
Sawka MN, Knowlton RO, Critz JB. Thermal and circulatory responses to repeated bouts of prolonged running. Med Sci Sports 1979;11:177-80. Sawka MN, Pimental NA, Pandolf KB. Thermoregula-
mun Chem Pathol Pharmacol 1976;13:713-22. Sedman AJ, Wilkinson PK, Sakmar E, Weidler Di, Wag-
ner JG. Food effects on absorption and metabolism of alcohol. J Stud Alcohol 1976;37:1197-214. Wilkinson PK, Sedman AJ, Sakmar E, Kay DR, Wagner JO. Pharmacokinetics of ethanol after oral administra-
tion in the fasting state. J Pharmacokinet Biopharm 1977;5:207-24.
Rangno RE, Kreeft JH, Sitar DS. Ethanol dosedependent elimination: Michaelis-Menten classical kinetic analysis. Br J Clin Pharmacol 1981;12:667-73.
Wilkinson PK, Sedman AJ, Sakmar E, Earhart RH,
CLIN PHARMACOL THER 1971;12:192-201.
tory responses to upper body exercise. EurJ Appl Physiol 1984;52:230-4. Sawka MN, Young AJ, Francesconi RP, Muza SR, Pandolf KB. Thermoregulatory and blood responses during exercise at graded hypohydration levels. J Appl Physiol 1985;59:1394-401. Mannering GJ, Deloria LB. The pharmacology and toxicology of the interferons: an overview. Annu Rev Pharmacol Toxicol 1986;26:480-7.
