Journal of Studies on Alcohol, Vol. 51, No.

I, 1990

Alcohol Consumption, Blood Alcohol Level and the

Relevance of Body Weight in Experimental Design
and Analysis*
M.S. DEVGUN a•m> J.A. DUNBARt

Department of Biochemistry, Law Hospital, Carluke, Lanarkshire ML8 5ER, Scotland

ABSTRACT. In healthy male subjects, intake of alcohol is significantly in people with different body weight; total body
not closely related to body weight although there is a signif- water is probably the key factor. This study raises doubts
icant correlation between consumption and body weight in about experimental designsand calculations of blood alcohol
the younger age group. No associationwas observedbetween whereby assumptionis made that correcting alcohol dose for
age and alcohol consumption. Age did show a significant body weight would give preciseinformation. Conclusionsfrom
association when expressed in terms of intake per weight. such studies could in fact be misleading. (J. Stud. Alcohol
Eight hours after consumption blood alcohol did not vary 51: 24-28, 1990)

T ISNOTPOSSIBLE
tocompare
results
ofstudies lean body mass. The information sought has obvious
that take into account consumption in terms of but very significant applications not only in forensic
grams of alcohol per kilogram body weight (Eisen- medicine, but also in the design of experimental
hofer et al., 1983; Freer and Statland, 1977a, 1977b; protocols involving alcohol studies in general.
Mikhailidis et al., 1983) with those of studies that
do not take body weight into account (Devgun et Method
al., 1985; Gill et al., 1982). Reasons for considering
A total of 53 apparently healthy male volunteers
body weight are usually not given, especially when
participated in the project. Their age range was 18-
no reference is made to the lean body mass or the
skin fold thickness as a measure of body fat. Volume 51 years, with a mean (+ SD) of 33 + 11 years.
Lunch at 12:30 PM and dinner at 5:30 PM were served
of distribution of alcohol is said to reflect the amount
of water present in various tissues of the body; this to all the subjectsand then the drinks (wines, beers,
spirits and soft drinks) were made available over a
would suggest that an obese person, with propor-
period of 5 hours starting at 7 •'M. The study was
tionately increased adipose tissue, should have a
designed to simulate the usual drinking practices of
higher blood alcohol level (BAL) when consuming
an alcohol dose similar to that taken by a lean
the locality. The drinks were served by an official
of the licensed premises, using standard measures for
person. That is, when there is proportionately less
dispensing drinks. Official records were kept of the
adipose tissue representing a percentage of body
number and type of drinks served. No alcohol con-
weight, then it is reasonable to assume that a rela-
sumption was allowed 12 hours before the experiment.
tively increased amount of body water would be
Subjects' body weights were also taken prior to lunch,
capable of holding larger amounts of alcohol.
but their previous drinking histories were not ob-
The object of our exercise was to investigate the
tained.
relationship between body mass, peak blood alcohol
Blood samples were taken before the drinks were
level and alcohol consumption but without relating
served and then at 1.5 and 8 hours after the 5-hour
these parameters to the skin fold thickness or the
drinking period. The 1.5-hour postdrinking blood
,

sample time was chosen becausethis is approximately

Received: July 27, 1987. Revision: September7, 1988. when average peak levels of alcohol are detected in
* Financial support from the Tayside Health Board is gratefully
the blood after a typical evening's drinking (Bowman
acknowledged.
• Dr. Dunbar is with the Department of Forensic Medicine, and Rand, 1980b; Saunders and Paton, 1981). All
Royal Infirmary, Dundee, Scotland. samples were analyzed for BAL using an enzymatic

24
 DEVGUN AND DUNBAR 25

TAm.E 1. Alcohol consumption(g) with regard to body weight (kg), age, total body alcohol (TBA) and blood alcohollevel (BAL, mg/dl)
at 1.5 hours (1) and 8 hours (2) after cessationof drinking. All values = mean + SD.

Age Weight n Consumption BAL TBAa

I 2 I 2

18-25 67 + 10 14 260 + 83 184 _+ 69 51 + 50 74 0.9

26-35 77 + 13 16 230 + 114 156 + 78 52 _+ 66 72 0.8
36-45 74 _+ 8 13 213 + 81 160 _+ 88 52 + 51 70 0.8
46-55 78 + 12 10 202 + 60 159 + 60 26 + 40 75 0.4

a Approximate TBA = 60% Body Weight x BAL (g/l).

method on ACA II (Automatic Clinical Analyzer, also presentedin Table 2 showing the elimination of
Du Pont Instruments) in a single batch-run. alcohol with respect to varying amounts of con-
sumption per kilogram of body weight. From Figure
Results 1 it is apparent that an increasein BAL is propor-
tional to consumption and that the metabolic rate
Table 1 shows alcohol consumption in relation to
parallels intake up to about 200 g (3 g/kg). Alcohol
age, body weight and BAL. Mean (+ SD) alcohol
consumption above 200 g (3 g/kg) appears to affect
consumptionwas 233 + 91 g (range 0-591), averaging
the rate of elimination of BAL very minimally.
3.3 g/kg (range 0-7.9). The mean weight of the group
Figures 2 and 3 show BAL in relation to con-
was 74 + 12 kg (range 55-100). The mean weight of
sumption and consumption per weight, respectively.
subjectsunder 25 years of age was significantly lower
No significant differences were observed with or
than that of the rest of the group (p < .05); this
without respect to body weight; the degree of asso-
group consumedlarger amounts of alcohol but with- ciation was similar at 1.5 hours after cessation of
out proportionately increased BAL at 8 hours after
drinking (r = 0.82, p < .005) but at 8 hours the
cessation of drinking. Consumption was negatively
association between BAL and consumption and BAL
correlated with body weight, however the degree of
and consumption per weight was r = 0.68 and 0.46,
association was best observed in the youngest age
respectively(p < .005). In general, the earliest detec-
group (r = -0.91, /9 < .005) in the study.
tion of significant amounts of blood alcohol at 1.5
However, BAL correlated with body weight only
hours after consumption corresponded to an intake
for ages below 25 (r = 0.55, p < .002) and above
of approximately 80 g (1.1 g/kg). At 8 hours after
45 years (r = 0.53,/9 < .002), both at 1.5 hours and
cessation of drinks, significant amounts of alcohol
at 8 hours after consumption (Table 1). Approxi-
correspondedwith ingestion of approximately 180 g
mation of total body alcohol (TBA) based on total
(2.2 g/kg). The figures also show that considerable
body water (TBW), approximately 60ø7obody weight
variations occur between alcohol consumption and
in men (Willatts, 1987), showed similar results for
BAL irrespective of body weight. For example, at
all age groups (Table 1). TBW was also calculated,
1.5 hours after ingestion (range 80 - 270 g) we can
allowing for age and weight, for each subject using
expect BAL as low as 40 mg/dL or as high as 250
the following equation (Watson et al., 1981):
mg/dL (Figure 2).
TBW = 20.03 - 0.1183 age + 0.3626 weight

The mean TBW was 43 + 4 liters with a range of T•m•.E2. Rate of eliminationof blood alcohol (BAL, mg/dL/h/kg)
based on kilogram body weight with varying amounts of alcohol
36-52. Age correlation with consumption in g/kg was
consumption (g/kg).
r = -0.38 (p < .01) but not at all significant with-
out reference to body weight. Table 1 shows that up
Alcoholconsumption BAL--rate of elimination
to age 45, regardless of level of consumption, BAL
Range Mean (ñ SD) Mean (+ SD)
at 8 hours postingestionwas not significantly different
in any of the subjects. 11 <2.0 1.44 _+ 0.46 NDa
Mean rate of alcohol elimination was 21 mg/dL/ 14 2.1-3.0 2.54 __ 0.31 0.20 _+ 0.06
13 3.1-4.0 3.55 + 0.25 0.32 _+ 0.02
h (range 17-23) between 1.5 and 8 hours after
13 4.1-5.0 4.94 _+ 0.34 0.32 + 0.0•
consumption. Mean disappearance rates (mg/dL/h) 2 >5.0 6.44 + 2.01 0.29 + 0.06
for different age groups were as follows: 18-25 = 23;
26-35 = 17; 36-45 = 21; and 46-55 = 23. Data are a ND = Not determined.
26 JOURNAL OF STUDIES ON ALCOHOL/JANUARY 1990

8-Oh

l'Sh

50
100- 101- 20(}
150 1õ1- 2,5,0
201- 30(}
2,51-4(•- >401 3oc -
aJeo_h•{cc•s•ml,t. ion , g
FIGURE1. Grams (g) of alcohol consumedand the rate (mean ñ
SD) of alcohol elimination (BAL, mg/dL/h). Blood alcohol
measurements taken at 1.5 hours and 8 hours after the end of drink-
ing period.
loc -

40• --

8-0 h 1-0 2-0 30 4-0 •0 •}0

alcohol consumption, g/kg
300-
FIGURE3. Relationshipbetweenalcohol consumptionwith respect
to body weight (g/kg) and blood alcohol level (BAL, mg/dL) at
ß

1.5 hours and 8 hours after the end of ingestion period.

200-

Discussion
100

It has been reported that the rate of oxidation of

alcohol is associated with BAL (Bowman and Rand,
1980b) but others report that it is related more to
the body weight than to BAL (Burgen and Mitchell,
l'5h
1972). The findings of this study are that BAL is
correlated with body weight for ages less than 25
and greater than 45 years only. This suggeststhat
- .J. TBW could be the key factor. TBW is higher in the
young (Bell et al., 1972) and therefore could allow
greater alcohol consumption in terms of body weight.
20O This would be reasonable if we can assume that
individuals aged 25-45 years have relatively greater
amounts of fatty tissue.
100
The approximated TBA was similar in all age
groups. From this it can be construed that TBW
I
does not vary significantly from second decade on-
16o 2oo 3o0 400 5oo 6oo wards; however, this would be presumptive as TBW
alcohol cons-mption, g (therefore, the volume of distribution) and percentage
Fioua• 2. Relationshipbetweengrams(g) of alcoholconsumedand
of body fat can vary significantly in people of same
blood alcohol level (BAL, mg/dL) at 1.5 hours and 8 hours after age and body weight (Bell et al., 1972; Diem and
the end of ingestion period. Lentner, 1970). Constancy of TBA at 8 hours after
 DEVGUN AND DUNBAR 27

consumption could result from increased rate of significance.In normal individuals variations in body
elimination via catalase and/or microsomal ethanol fat (approximately 14ø70body weight) can be up to
oxidizing system (MEOS) found in the endoplasmic 10ø70(Bell et al., 1972); but in those with abnormal
reticulum; especially in subjects with excessivecon- body weight, proportion of body fat could range
sumption. This would be appropriate as Km for from 8-50ø7o(Willatts, 1987). Therefore, giving a dose
MEOS is greater than the K m of the metabolic of alcohol based on body weight would not give
pathway involving alcohol dehydrogenase (Lieher, better controlled trials or make the interpretation of
1984); and consistent with our results in Figure 1. results any easierin this important field of forensics,
Role of stress-associatedrise in adrenaline ("swift science and medicine. Results from studies can be
increase") and the increased rate of elimination of compared with one another only if protocols adopted
BAL to maintain TBA constancy may also be sig- for experimentation are similar in design; that is, if
nificant (Yuki and Thurman, 1980). alcohol dose has been considered on the basis of
The association of age with consumption, with total body water.
respectto weight only (i.e., g/kg), suggeststhat body
weight increaseswith age and that alcohol consump- References
tion is moderated, resulting in similar levels of BAL.
Perhaps this is related to age-associatedincrease of BELL, G.H., DAVIDSON, J.N. AND EMSLIE-SMITH,D. Water, elec-
fatty tissue, decreasein TBW (Bell et al., 1972) and trolyte and acid-base metabolism. In: BELL, G.H., DAVIDSON,
to drinking behavior (i.e., time taken to consume a J.N. AND EMSLm-SMI?H,D. Textbook of Physiology and Bio-
given amount of alcohol). The results in Table 1 chemistry, 8th Edition, Baltimore: Williams & Wilkins Co.,
1972, pp. 671-704.
show that BAL was similar in all age groups (up to
BOWMAN,W.C. AND RAND, M.J. Aliphatic alcohols:Ethanol. In:
the age of 45) at 8 hours postingestion. This is BOWMAN,W.C. AND RAND, M.J. Textbook of Pharmacology,
suggestive of a homeostatic type mechanism that 2d Edition, Boston, Mass.: Blackwell Scientific Pubns, Inc.,
controls BAL, matching increasingload by accelerated 1980a, 8.12-8.16.
elimination and minimizing the elevation in BAL. BOWMAN,W.C. AND RAND, M.J. Social pharmacology: Drug use
for nonmedicinal purposes--ethanol. In: BOWEN, W.C. AND
This is reflected by the increasedrates of elimination
RAND, M.J. Textbook of Pharmacology, 2d Edition, Boston,
found in this study; these values are higher than Mass.: Blackwell Scientific Pubns, Inc., 1980b, 42.2-42.24.
those reported elsewhere (Bowman and Rand, 1980a; BURGEN, A.S.V. AND MITCHELL, J.F. Central nervous system
Gee, 1979). Differences may be due to food, sex, depressant II. In: BUROEN,A.S.V. AND MITCI-IELL, J.F. Gad-
age, race, type of alcoholic beverage and the levels dum's Pharmacology, New York: Oxford Univ. Press, Inc.,
1972, pp. 25-38.
of consumption. Our subjects were given a meal
DEVOUN, M.S., DUNBAR, J.A., HAOART, J., OOSTON, S.A. AND
before the experiment; this could account in part for MARTIN, B.T. Biochemical measurements and screening for
higher levels of consumption and for the body's alcohol abuse in drunk drivers. In: KAYE, S. AND MEmR, G.W.
ability to handle increased blood alcohol levels. How- (Eds.) Alcohol, Drugs and Traffic Safety. National Highway
ever, such a controlling mechanism may or may not Traffic Safety Administration, DOT HS 806 814, Washington:
Government Printing Office, September 1985, pp. 345-350.
be present in alcoholic subjects or in other disease
DIEM, K. AND LENTNER, C. (Eds.) Composition of the body. In:
states.
Documenta Geigy: Scientific Tables, 7th Edition, Basel: Ciba-
Consumption exceeding 200 g (3 g/kg) seemed to Geigy Ltd., 1970, pp. 517-522.
alter the metabolic rate minimally. Above this point EISENHOFER,G., LAMBIE, D.G. AND JOHNSON, R.H. Effects of
the limiting factor in ethanol metabolism could be ethanol on plasma catecholaminesand norepinephrineclearance.
Clin. Pharmacol. Ther. 34: 143-147, 1983.
the availability of coenzyme NAD+. An increased
FREER,D.E. AND SWATLAND, B.E. The effects of ethanol (0.75 g/
pharmacological tolerance to alcohol would then be kg body weight) on the activities of selected enzymes in sera
expectedin a person with 70 kg body weight with a of healthy young adults. I. Intermediate-term effects. Clin.
regular intake of over 200 g. Our subjects showed Chem. 23: 830-834, 1977a.
considerable variations in their ability to consume FREER, D.E. AND STATEAND,B.E. The effects of ethanol (0.75 g/
kg body weight) on the activities of selected enzymes in sera
alcohol and in the consequential BAL results. This
of healthy young adults. H. Interindividual variations in re-
is a reflection of pharmacological as well as behav- sponseof .y-glutamyltransferaseto repeated ethanol challenges.
ioral tolerance to alcohol. Clin. Chem. 23: 2099-2102, 1977b.
However, it is possible body fat/water has signif- GEE, D.J. Lecture Notes on Forensic Medicine, 3d Edition, Boston,
icance in the metabolism of alcohol either as a storage Mass.: Blackwell Scientific Pubns, Inc., 1979, pp. 82-83.
depot or via changes in ethanol concentration and GILL, G.V., BAYElS, P.H., FLEAR, C.T.G., SKILLEN, A.W. AND
DIOOLE, P.H. Acute biochemical responses to moderate beer
therefore the pharmacokinetics. Our data suggest
drinking. Brit. Med. J. 285: 1770-1773, 1982.
body weight is probably not a significant factor; as LmBER, C.S. Metabolism and metabolic effects of alcohol. Med.
alcohol is iso-osmoticallydistributed in the body, it Clin. N. Amer. 68: 3-31, 1984.
is the volume of distribution (i.e., TBW) that is of MIKHAn.IDIS, D.P., JEREMY,J.Y., BARRADAS,M.A., GREEN, N.
28 JOURNAL OF STUDIES ON ALCOHOL/JANUARY 1990

AND DANDONA,P. Effect of ethanol on vascular prostacyclin WILLATTS,S.M. Normal water balance and body fluid compart-
(prostaglandin I:) synthesis, platelet aggregation, and platelet ments. In: W•i•ATTS, S.M. Lecture Notes on Fluid and Elec-
thromboxane release. Brit. Med. J. :z$7: 1495-1498, 1983. trolyte Balance, 2d Edition, Boston, Mass.: Blackwell Scientific
SAUNDERS, J.B. AND PATON,A. Alcohol in the body. Brit. Med. Pubns, Inc., 1987, pp. 3-31.
J. 283: 1380-1381, 1981. YLrKI, T. AND Tn•URMAN, R.G. The swift increase in alcohol
WATSON,P.E., WATSON,I.D. AND BATT, R.D. Prediction of blood metabolism:Time course for the increasein hepatic oxygen
alcohol concentrationsin human subjects:Updating the Wid- uptake and the involvement of glycolysis. Biochem. J. 156:
mark equation. J. Stud. Alcohol 4:z: 547-556, 1981. 119-126, 1980.