NORMAL VARIATIONS IN TOTAL KETONE BODIES IN SERUM
AND URINE OF HEALTHY YOUNG MEN. By R. E. JOHNSON,
F. SARGENT, II and R. PASSMORE. From the Departments of
Physiology, University of Illinois, Urbana, U.S.A. and Edinburgh
University.
(Received for publication 22nd March 1958)
To establish normal values and ranges for studies of ketosis, total ketone
bodies have been estimated in the urine and serum of 208 healthy young men,
eating non-ketogenic adequate diets and engaged in moderate daily physical
activity. Under standardized post-absorptive conditions timed specimens of
urine were collected, and venous blood was drawn. The urinary excretion rate
of ketone bodies differed significantly in summer and winter. In hot weather
the mean rate was 0 9 ,c-mole/min.; in cold weather it was 2-8 ju-mole/min.; in
cool weather it was 1X7 ,c-mole/min. For serum, regardless of season, the mean
was 0 7 m-mole/l. If two standard deviations be taken as the normal range, it
is suggested that for most conditions the upper limit of normal for serum be
considered as 1-4 m-mole/l. and for urine 5 u-mole/min.
AN interest in ketosis which may arise in healthy men from nutritional causes,
especially starvation, and after prolonged exercise (the Courtice-Douglas
effect), led us to inquire: "What are the normal values for total ketone bodies
in blood and urine? " A systematic study of changes in ketone metabolism
in man demands knowledge of the normal, yet present information is scanty
and unsatisfactory. For instance, the Handbook of Biological Data [Albritton,
1953] has no separate lists for the individual ketone bodies in urine, and only
gives the total excretion in daily specimens collected at random without
regard to previous dietary history or activity of the subjects. Furthermore,
analytical methods have been unsatisfactory and tedious. For betahydroxybutyric
acid, which may account for half or more of the total ketone
bodies, no qualitative test exists, and hitherto many quantitative methods
have not given true values for it.
During metabolic studies with 208 subjects, we have been able to collect
specimens of blood and urine under standardized post-absorptive conditions
and to analyze these with satisfactory quantitative and qualitative methods.
Knowledge of the subjects' previous dietary history, daily activity, and
environment was adequate. Hence, we are able to assert that for healthy
young men we know the normal range of concentration of total ketone bodies
in serum; the normal rate of excretion of these substances in urine under
post-absorptive conditions; and the relation between quantitative and
qualitative examination of urine collected under these conditions. Qualitative
results have been reported by Sargent and Consolazio [1950] which support
VOL. xLIII, No. 4.-1958 339 23
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
Johnson, Sargent and Passmore
the view that exposure to cold increases the rate of excretion of acetone plus
acetoacetic acid. We have confirmed this observation quantitatively for
total ketone bodies.
METHODS
Subjects.-Three studies have been conducted: summer, 1955 (Indiana); winter,
1954 (Wisconsin); and winter, 1957 (Edinburgh). In each of the first two, ninetynine
healthy young airmen were the subjects, and in the last, ten healthy young
medical students.
Activity and Environment.-Daily work was moderate, consisting of walking
outdoors to and from classes, other duties and meals. At most other times the
subjects were living a sedentary life and were not much exposed to the weather,
even though the outdoor temperatures were rather severe.
TABLE I.-SUBJECTS AND REGIMENS
Groups of Subjects Duration of Total Percentage of Calories from
subjects number regimen Calories
days Cal/day Protein Fat Carbohydrate
Summer, 1955, Indiana
A* 88 14 3400 15 34 51
Bt 11 14 4100 14 41 45
Winter, 1954, Wisconsin
C* 87 14 3200 14 33 53
Dt 12 14 4500 15 40 45
Winter, 1957, Edinburgh
Elt10 2 3000 15 45 40
* This diet consisted of canned items in large variety with fresh bread.
t This diet consisted of abundant amounts of fresh and frozen foods in wide variety.
$ This diet consisted of dried milk solids with hydrogenated vegetable fat and carbohydrate
added.
Diet.-This is described summarily in Table I. The subjects were under continuous
nutritional observation. It was intended that the amounts of the proximate
principles of their diets should be representative of usual North American and
British practice and should be adequate in quality. Unlimited fluids were provided.
In the two American experiments the daily physical activity of all subjects was the
same. Groups B and D received a luxury diet of the best procurable natural foods.
Collection of Urine and Blood.-On the night before the observation food was
last eaten at about 8 P.M. In the morning 100-200 ml. of water was drunk and the
subjects came, without food, to the laboratory. In the 1954 and 1955 experiments
the subjects emptied their bladders and then lay down for 2 hr. without smoking,
drinking or eating. During this period a specimen of urine, timed to the nearest
minute, was collected and its volume measured to the nearest ml. Venous blood
was drawn at the midpoint. In the 1957 experiments the subjects were ambulatory
and no blood was drawn.
Quantitative Analysis.-There is as yet no rapid and accurate quantitative
method for determining total ketone bodies. In particular, the oxidation of betahydroxybutyric
acid to acetoacetic acid by dichromate at 100° C. has usually been
reported as only 60-80 per cent complete. We have succeeded in making this yield
100 per cent by using temperatures over 1000 C. together with metaphosphoric acid,
which catalyzes the oxidation. This substance has the further advantage of being an
excellent protein precipitant.
340
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
Ketones in Serum and Urine
For total ketone bodies, 10 ml. of 0 09 M-metaphosphoric acid are placed in a
cone-tipped centrifuge tube. To this is added 0 5 ml. of serum or urine, and the
whole is mixed by inversion. After centrifuging, 3 ml. of the supernatant are placed
in an ampoule (or, rather better, in a screw-topped tube fitted with a teflon plug) and
then 1 ml. of 3 N-sulphuric acid containing potassium dichromate in 0-02 M concentration
is added. The ampoule or tube is sealed and autoclaved for 1 hr. at
15 lb. pressure. After cooling, the mixture is analyzed by the method of Michaels,
Morgan, Liebert and Kinsell [1951] or by the method of Thin and Robertson [1952].
The first relies on the formation of the 2: 4 dinitrophenylhydrazone of acetone, which
is extracted into carbon tetrachloride; in the second acetone diffuses in a
Conway plate into alkaline salicylaldehyde. In both methods the intensity of
the final colour is measured. Reagent blanks and standards of calcium zinc betahydroxybutyrate
are included in each batch of samples.
Qualitative Analysi8.-The method of Rothera [1908] has been standardized to
give maximum sensitivity and uniform comparability from time to time. About
5 g. of anhydrous ammonium sulphate is added to a clean dry test tube; there must
be enough to saturate the final solution and leave an undissolved excess. In succession
and with mixing after successive additions, 0.5 ml. of urine, 0-2 ml. of freshly prepared
0 07 M-sodium nitroprusside, and 0-2 ml. of 5-N-ammonium hydroxide are added.
Colour is recorded as 0-4 plus at 2, 5 and 10 min., and the maximum reading is
accepted.
RESULTS
Quantitative: Serum.-Between winter and summer there were no significant
differences in concentrations of total ketone bodies in the serum
(Table II), and no significant differences between groups A and C and groups
TABLE II.-CONCENTRATIONS OF TOTAL KETONE BODIES IN SERUM
Group of Observations Mean Standard Coefficient Outdoor
subjects number m-mole/l. deviation of variation temperature m-mole/l. per cent ° C.
Summer, 1955, Indiana
A 88 0-64 0.20 31 +26 to +27
B 11 0-68 0-17 25 +26to +27
Winter, 1954, Wisconsin
C 87 0-75 0-19 25 -7 to +7
D 12 0-96 0-38 39 -7to +7
Weighted means, Indiana plus Wisconsin
AC 175 0-69 0.20 29
BD 23 0-83 0.28 34
all together 198 0-71 0X21 30
Statistical analysis: by students' "t" test, no differences between groups or seasons
were significant at a probability level of 5 per cent.
B and D. Therefore, a weighted mean for all subjects is justified. For
all 198 specimens the mean was 0-71 m-mole/l., the standard deviation
0-21 m-mole/l. and the coefficient of variation 30 per cent.
Quantitative: Urine.-The urinary excretion of total ketone bodies
varied significantly with season (Table III). In severe winter cold excretion
was greatest; in warm summer weather, least; and in the moderate winter
341
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
Johnson, Sargent and Passmore
weather of Scotland, intermediate. In all groups, the coefficient of variation
was larger than for serum. No weighted means would be statistically
permissible for urine.
TABLE III.-URINARY EXCRETION OF TOTAL KETONE BODIES
Standard Coefficient Outdoor Groupjeso O bservationseMean deviation of variation temperature sm-mole/min. per cent 0 C.
Summer, 1955, Indiana
A 88 0-80 0 30 38 +26 to +27
B 11 1O08 0585 79 +26 to +27
Winter, 1954, Wisconsin
C 87 2-56 1-28 50 -7to +7
D 12 3-10 1*32 43 -7 to +7
Winter, 1957, Edinburgh
E 20 1-73 0-94 54 +8 to +14
Statistical analysis: by students' "t" test, differences were not significant between
groups within summer and within winter at the 5 per cent level. All differences between
A and B groups and C and D groups were significant at the 1 per cent level or below.
Differences between the E groups and all AB and CD groups were significant at the
5 per cent level or below.
TABLE IV.-ROTHERA REACTION IN URINE
Observations Strength of reaction
Groups nubservaln
Groups number 0 tr 1 + 2 + 3 + 4 +
Summer, 1955, Indiana
A 170 169 1 ..
B 24 24 .. ..
Winter, 1954, Wisconsin
C 158 116 18 15 4 3 2
D 22 13 3 3 2 .. 1
Winter, 1957, Edinburgh
E 20 19 1 ..
Statistical analysis: by the "Chi Square" test, there was no difference between A
and C or B and D groups within summer or within winter, significant at the 5 per cent
level. There was a difference between both winter 1954 groups, and all summer 1955
and winter 1957 groups, significant at the 0-01 per cent level.
Qualitative: Urine.-There was a significant difference between severe
winter weather and summer or mild winter weather (Table IV). Among the
394 specimens tested, positive reactions of + 1 or above were obtained only
in Wisconsin. These qualitative findings for acetone plus acetoacetic acid
are in agreement with the quantitative results of Table III for total ketone
bodies; and they confirm the observations of Sargent and Consolazio [1950]
for conditions of extreme cold.
342
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
Ketones in Serum and Urine
DISCUSSION
The interpretation of means and ranges is always a matter of statistical
judgment. For total ketone bodies in serum, we suggest that twice the
standard deviation be considered the normal range. Our groups B and D
had the higher mean figure. If this is taken and two standard deviations
added, the upper limit of normal for total ketone bodies in serum is
1-4 m-mole/l., regardless of season. The existence of a correlation with
temperature complicates the statistics for urinary excretion. However, as
a general rule we should place 5 ,u-mole/min. as the upper limit of normal,
realizing that it will be less in warm weather. Recalculation of data in the
Handbook of Biological Data [Table 116, Albritton, 1953] gave an average of
2-3 ju-mole/min., and an upper limit of 5 7, for daily specimens collected
without control of diet or environment.
For physiological studies of ketone metabolism we urge the use of units
of chemical equivalence, reform of inaccurate terminology, and standardization
of conditions for collecting urine and blood.
It has been common for investigators to express their results in terms of
mg. of acetone per 100 ml. of urine or blood. This can lead to confused
thinking. Table V demonstrates that, if one considers only concentration by
weight in the blood, glucose would appear to be the predominant carbohydrate
TABLE V.-SOME ILLUSTRATIVE CONCENTRATIONS OF INTERMEDIARY
METABOLITES IN BLOOD
Metabolite Metabolite CCoonnddiittiioonnss mBg.y/1w0ei0gmhlt. Beyquicvhaelmeinccael
m-mole/l.
Glucose . . . At rest 81 4.5
Free fatty acid . . At rest 28 1.0
Lactic acid . . At rest 9 1-0
Lactic acid . . After exercise 81 9.0
Pyruvic acid . . At rest 2 0-23
Pyruvic acid . . In severe beriberi 10 1.1
Ketone bodies . . At rest 5 0.8
In nutritional ketosis 23 4.0
(or derivative), and total fatty acid the predominant fat (or derivative). As
judged by equivalent chemical units, however, such is certainly not always
the case. In nutritional ketosis the molar concentration of total ketone
bodies may equal that of glucose and far exceed that of free fatty acids;
and after anaerobic work of short duration, the molar concentration of lactic
acid is often at least double that of glucose. Even metabolites that are
normally present in very small concentrations, like pyruvic acid, under some
circumstances may become quantitatively important.
Terminology for discussions of ketosis needs modification. The terms
"ketonuria " and "ketonaemia" were originally given a pathological significance,
because they were based on qualitative tests which could not detect
small concentrations. However, modern quantitative methods show that
there are always measurable concentrations of ketone bodies in urine and
343
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
Johnson, Sargent and Passmore
blood. Therefore, the pathological connotation might better be served by the
terms "hyperketonuria" and "hyperketonsemia".
We recommend strongly the routine use of timed specimens of urine
collected when the subjects are post-absorptive, and blood drawn during the
same period. The two advantages that would accrue would be comparability
of data from laboratory to laboratory, and applicability of data to calculation
of renal parameters. In view of the correlation between temperature and
renal excretion of total ketone bodies, we recommend control of the environment,
or, if that is not possible, recording of environmental conditions under
which the study may be conducted.
ACKNOWLEDGMENTS
All quantitative analyses were performed by Mrs. Evelyn Robbins, Mrs. Laura
Sawyer and Mr. David Shirling, whose skill and industry we acknowledge here.
Qualitative analysis was done by R. E. Johnson and N. Sperelakis.
This investigation was supported in part by contract AF 18(600)-80 between the
U.S. Air Force and the University of Illinois; and in part under a grant to R. Passmore
from the Medical Research Council.
In 1957-1958 R. E. Johnson was on sabbatical leave and the recipient of a U.S.
National Science Foundation Senior Post-doctoral Research Fellowship for study at
Edinburgh University. F. Sargent, II, was on sabbatical leave and the recipient of a
John Simon Guggenheim Memorial Foundation Fellowship for study at Oxford
University.
REFERENCES
ALBRITTON, H. (editor) (1953). Handbook of Biological Data. Washington, D.C.:
National Research Council.
MICHAELS, G. D., MORGAN, S., LIEBERT, G. and KINSELL, L. W. (1951). "Studies in
fat metabolism. I. The colorimetric determination of ketone bodies in biological
fluids", J. clin. Invest. 30, 1483-1485.
ROTHERA, A. C. H. (1908). "Note on the sodium nitroprusside reaction for acetone",
J. Physiol. 37, 491-494.
SARGENT, F., II and CONSOLAZIO, C. F. (1951). "Stress and ketone body metabolism",
Science, 113, 631-633.
THIN, C. and ROBERTSON, A. (1952). "The estimation of acetone bodies", Biochem. J.
51, 218-223.
344
Downloaded from Exp Physiol (ep.physoc.org) by guest on December 27, 2011
கருத்துகள் இல்லை:
கருத்துரையிடுக