Download - QUANTITATIVE

QUANTITATIVE RELATIONSHIPS BETWEENBLOODANDURINEKETONELEVELS IN DIABETIC KETOSIS

By HELEN EASTMANMARTIN AND ARNE N. WICK(From the Departments of Physiology and Medicine of the University of Southern California,

and the Department of Medicine of the Los Angeles General Hospital, Los Angeles,and the Scripps Metabolic Clinic, La Jolla, California)

(Received for publication September 17, 1942)

In the study of ketosis in diabetic patients, cer-tain problems have gone largely unanswered dueto the lack of sufficiently accurate methods forblood ketone determinations. With the develop-ment of blood ketone methods accurate at low aswell as high blood concentrations, it has beenpossible to study the problems of (1) the renalthreshold for the ketone bodies, and (2) the quan-titative relationship between the urinary ketoneexcretion and the blood ketone level.

The studies to be presented summarize the re-sults obtained on (1) 13 diabetic patients in mildketosis, produced by withdrawing insulin for 12 to24 hours, a ketogenic diet for 12 to 24 hours, ora combination of a ketogenic diet and insulin with-drawal; and (2) 7 patients entering the LosAngeles County Hospital in moderate to severeketosis (several in diabetic "coma"). In bothtypes of patients, after initial levels of blood andurine ketone bodies were obtained, insulin andfluids were given as required by .the individualcase, and blood and urine ketone levels were. fol-lowed until ketosis, as indicated by the qualitativeurinary acetone test (modified Lange test), dis-appeared. The blood ketones 1 were determinedby the Barnes-Wick (1) method, and the urineketones by the Van Slyke (2) gravimetric methodin most instances.

Table I summarizes the results obtained in thediabetic patients in mild ketosis, and Table II theresults in patients -in moderate to severe ketosis.Figures 1 and 2 give the clinical course, duringtherapy, of two of the most completely studiedcases, one in mild ketosis, the other in severeketosis. These are patients "Ha" and "Wi,"from Tables I and II, respectively. The figuresillustrate how the data listed in the tables wereobtained.

1 Total blood and urine ketones calculated as beta-hy-droxybutyric acid (1).

From the analysis of the data given in thesetables and figures, certain facts are apparent. InTable I (record of patients with blood ketonelevels between 1 and 30 mgm. per cent), it isseen that even with very low blood ketone levels(5 to 10 mgm. per cent) small amounts of theketone bodies appeared in the urine. This in-dicates that the absolute renal threshold for theketone bodies in diabetic patients is very low-under 10 mgm. per cent-in the majority of cases.While it is evident that the threshold for theketone bodies is very low in diabetic patients,relatively small amounts appeared in the urine,less than 100 mgm. per hour, until blood levelsover 20 mgm. per cent are reached. This givesa ratio under 10 for

urine excretion mgm. per hourblood ketones mgms. per cent

in 10 of the patients listed in Table I. In onepatient ("P"), all the ratios were under 10, withone exception, where there was a marked increasein urine output per hour. In two patients ("Ho"and "St"), there were much higher ratios. Cer-tain variations in the ratios at low blood levelscan be explained on the basis of marked variationsin urinary output per hour.

The great increase in urinary ketone excretionper hour with blood ketone levels over 20 mgm.per cent is seen in Table II (blood ketone levelsfrom 35 to 150 + mgm. per cent). The amountof urine ketones excreted per hour at comparableblood levels varied greatly from patient to patient.As the amount of the urinary ketone excretionis very small as contrasted to the amount utilized(3), this may partially explain this difference.These results in patients in moderate to severeketosis indicate a complex relationship betweenthe blood ketone level and the urinary ketone out-put per hour. When the urinary ketone outputper hour is charted against the blood level, a

235

HELEN EASTMANMARTIN AND ARNE N. WICK

TABLE I

Keto excretin in mild ketosis

Patient | Time | Urine volume | Urin ketones*) Urine ketones mgm. per hour Blood ketonesPatient Time Urine volume Urine keton~ Blood ketones mgm. per cent

3.43.51.61.0

7.44.5

22.44.5

5.95.1

3.71713

7.9

5.72.1

0

1.120.832.72.63.2

5.87.9

. 0

6.80

0

0

0

0.81.83.61.590

46.20

3.42018.5

7.5

mgm. per cent

3016.415.512.8

1914.7 to 21

5.44.1

1642

21 to 15.8 to 6.33.6 to 2.42.4 to 3.63.6 to 7.2

5.77.9

8 to 1.2

12.85.8 to 62.9 to 2.21.7 to 3.13.1 to 0.7

8.88.1

4.7 to 2.4

8.25.32.71.4

7.8 to 3.8

6.06.6

10.02.51.7

5.14.8

4.31.20.71.0

* Calculated for time period listed.t Urine ketones fractionated into beta-hydroxybutyric acid and acetone. Amount listed represents beta-hydroxy-

butyric acid.t Urine ketones represent beta-hydroxybutyric acid. (Mercury precipitate dissolved, distilled and reprecipitated.)

hyperbolic type of curve is obtained (Figure 3-

patient "St"-Table II). When these data fromFigure 3 are expressed as a relationship betweenblood level and a ratio between urine ketone

excretion mgm. per hour and blood ketone levelin mgm. per cent, a straight line is obtained, withthe point of origin passing through the blood levelaround 20 mgm. per cent (Figure 4). In the

1. "Ha"t

2. "P"t

3. "Du"t

4. "Ho"

5. " Str "

6. "Hal"

7. "D"t

8. "Si"

9. "N"

10. "Q"

11. "B"t

12. "Be"

13. "St "t

minutes97864352

1661925540

201169

269537065

5560

315

5760576651

60,6090

70606060

175

5662645643

68180

60605862

cc. per hour414

1,1501,310

559

123106933630

12841

118386283328

71276

210

98110697494

38138504

10718016240

63

115420430573725

424255

350370671203

mgm. pe hour101582512.9

14180.5

12118.3

9410.2

63.7523942.5

32.516.4

0

14.44.96.96.36.1

5164

0

55.9000

0

4.8111.835.7

3.960

-31.50

14.62413

7.5

236

BLOODAND URINE KETONELEVELS

C

UrinelKetonesMgmo/hr.

125

100

75

So

25

ourse of mild Ketosis. with therapVBloodKetones

Mgm.-0-Z -o

0 1 2 3 4 5 6Durarion of Ketosis in hours

FIG. 1. THE RAPID FALL OF THE BLOODKETONELmLANDTHE SMALLURINARY KETONEOUTPUTAT Low BuODKETONELEvELs (Patient "Ha," Table I)

majority of the cases, a somewhat similar relation-ship was found, with a marked falling off of theratio at blood levels under 20 mgm. per cent.

Studies in 5 patients included the determinationof both acetone (including acetoacetic acid andacetone) and beta-hydroxybutyric acid in theurine, at varying blood levels-2 to 78 mgm. per

cent. The results are listed in Table III. Inpatients "Ha" and "K," acetone represented ap-

proximately 80 per cent of the ketone bodies inthe urine at the lower blood levels. In patient"P," the urinary acetone rose from 11 per cent,at a blood ketone level of 19 mgm. per cent, to 41per cent, at a blood ketone level of 2.9 mgm. per

cent. In patient "B," the urinary acetone rose

from 42 to 100 per cent, with correspondingchange in blood ketone level from 6 to 1.7 mgm.per cent. In patient "D," there was only a slightincrease in percentage of acetone at the lowerblood ketone levels.

In studying the data summarized above, certainother points of interest were noted. In all thepatients in moderate to severe ketosis, the carbondioxide combining power was determined re-

peatedly and synchronously with the blood ketonedeterminations. The lack of specific correlationin different patients between the carbon dioxidecombining power and the blood ketone level is

seen in Table IV. (This reaffirms the previouswork of many.) In this table it is seen that. onepatient with a carbon dioxide combining power of10 volumes per cent had a blood ketone level of80 mgm. per cent, while another patient with acarbon dioxide combining power of 16 volumesper cent had a blood ketone level over 150 mgm.per cent. Since the carbon dioxide combiningpower is only a measure of the alkali reserve, andthis is determined by many factors other than thedegree of ketosis in uncontrolled diabetes, it isobvious why the two determinations are not moreclosely related.

Table V shows the low urinary ketone outputper hour in a patient with uremia and oliguria.In this patient with blood ketone levels over 150mgm. per cent, 61.7 to 455 mgm. per hour wereexcreted in the urine.

DISCUSSION

There are few quantitative figures given in theliterature on the relationship between blood andurine ketone levels. The impression,. however,is given that there is no correlation. As Petersand Van Slyke (4) state: "The relation betweenblood and urine ketones has not been studied;but the data of Allen, Stillman and Fitz show apeculiar lack of association between the two."In studying the data of these authors (5), it isnoted that the total urinary ketone excretion for24 hours, with varying diets or starvation, iscorrelated apparently with one or two bloodketone levels. As variations occur during the 24hours, it is impossible to draw conclusions fromtheir data, although their results do show verysmall urinary ketone output at low blood ketonelevels. Also there are very few references to therenal threshold for the ketone bodies in the litera-ture. The work of Wilder (6) on the injectionof beta-hydroxybutyric and acetoacetic acid intodogs, suggested a urinary threshold, since 0.4gram per kgm. of body weight had to be injectedbefore beta-hydroxybutyric acid appeared in theurine. Briggs and Shaffer (7) conclude fromtheir studies that acetone is a non-threshold sub-stance, and the concentration in the urine parallelsthe concentration in the blood. These authors,and also Widmark (8), state that under certainconditions acetone can diffuse into the urinarybladder directly from the blood, like alcohol, with-

237


TABLE II

Ketone excretion in moderate to severe ketosis

Urine Total urine Urine Urine ketones mgm. per hour BloodPatient Time volume ketones ketones Blood ketones mgm. per cent ketones

mgm. per ceei80 to 86

7575746016.4

12.81.3 to 2.32.3 to 5.2

60 to 41.5, av. 50.741.5

29.9 to 22.9, av. 2623.118.316.2

786651.738

36.929.924.3

11.1

10

64

25

4

36.326.9 to 25.822.4 to 23.5

5.1

108 to 150+69

* Calculated for time period listed.t Some urine lost.t Urine ketones fractionated into acetone and beta-hydroxybutyric acid. Amount listed represents beta-hydroxy-

butyric acid. § Patient in uremia.

out excretion through the kidneys. Widmarkfurther claims that for acetoacetic acid there is no

correlation between the blood and urine curves,

indicating secretory activity of the kidney. Allthis early work, however, is open to the criticismof relatively inaccurate blood methods. Shipleyand Long (9) in recent years have establishedevidence that in the rat, there is a definite renalthreshold-between 25 and 30 mgm. per cent.

Our results give definite quantitative data con-

cerning the renal threshold for the ketone bodies,and the correlation between blood level and theurinary excretion per hour. The urinary excre-

tion of ketone bodies varies with the blood level,but does not act similarly to such substances as

urea. When the urinary ketone output per houris charted against the blood level, a somewhathyperbolic type of curve is obtained, with a

1. "XWi"

2. "St"

3. "K"t

4. d"B "

5. "4Go"

6. "W"

7. "H"§

minutes24025016526022519017018017085

210

290110180180175125

24065655760

11013012012060

1305060

375140

155135115120120

185135

cc. per hour589444336426613805t380282t405494128

370327297

-i160(?)284198

27255

250137200

292180285365110

128516610138178

170t -

131t82

12121t

11753

mgm.22,05018,120

9,00012,560

9,9504,630+1,135

955+340

0136

17,3052,3401,322

534343

77

3,01456

1169

12

485376237

00

2,710922397311

40

618+158+98

00

977330

mgm. per hour*5,5124,3503,2702,9002,6501,400+

400318+120

039

3,5801,270

440=178

11737

75452

1079

12

264174119

00

1,2541,106

3975017

239+70+51

00

316147

66.55843.639.244.285.4

9.40

10.4

70.630.616.8i7.7

6.42.3

9.70.82.10.24

7.25.84.900

19.6

15.934

6.6+2.7+2.200

I I-

I I I

238


Course of. severe Ketoes, with therapy9Urine Ketones'in 119mi/hr.

Blood KetonuMom%

5000

4000

;000o

2000

1000

0 4 12 16 ZO 24Duration of Ketosis in hours

28 32

40

60

Re lation sh ip of UJrinec excretion of Ketoe1nsto blood level In severe Ketosis

S~~~~~~

J~~~~~~~~~~~1 '';

FIG. 2. THE PROLONGEDPERIOD OF SEvERE KETOsIS (20 HOURS) WITH THE VERY LARGE URINARYKETONEOUTPUTAT THE HIGH BLOODKEONELEVELS (Patient "Wi," Table II)

Relationship of Urine excretion of Ketonesto blood level in severe Ketosis

Ketones

Mgm%60

40

0

,oI-

10 20 10 40 SOUrine excrerion Ketones-Magm/he

Blood Ketones- Mgm-

50 70

1000 2000 3000 ao0Urine Ketones Mgni./hr.

FIG. 3. The INCREASE IN URINARY KETONEOUTPUTPER HoURWITH RISING BLOODKETONELEVms (Patient"St," Table II)

FIG. 4. EXPRESSING THE DATA FROMFIGURE 3 AS A

RATIo BETWEENBLOOD KrONE LEVEL AND THE RATioBETWuNURINE KETONESPER HoURAND BLOODLEVEL,A STRAIGHT LINE RELATIONSHIP IS OTAMNED(Patient"St," Table II)

BlodKetonesMgan

60

50

40

301~

20

10

239


TABLE III

Fractionation of urine kdones into bde-hydroxybuyricacid and aceon*

TotalTotal beta- Beta-

Patientacetone hydroxy- Ace hydrowy- BloodPtetin urine butyric tone butyric ketones*in urine

mgm. mgm. Per cext Per cen mgm. per cen1. "B" 3.1 4.3 42 58

2.3 4.5 33 67 69.6 12.2 44 56 6.6

18 38.2 32 68 103.4 3.7 48 52 2.56.8 0 100 0 1.7

2. "K" 34 68 34 6635 385 8 92 78

460 2,300t 17 83 7584 *329 20 80 7834 56 37 63 66

113 116 50 50 5138 9 81 19 3868 12 85 15

3. "Ha" 80 163 33 67 3084 76 53 4747 7.1 87 13 16.465 17.9 78 22 15.541.7 17.2 79 21 12.878 28.9 73 27. 0.5

4. "P" 48.9 391 11 89 1946.9 258 15.4 84.6 21 to 14.718 85 17.4 82.6 5.4

9.5 26.2 27 73 4.18.4 12.2 41 59 2.9

S."D" 23.4 270 8 9248.1 314 13.3 86.7 16

5.3 28.8 15.5 84.5 2

* Acetone represents acetoacetic acid and acetone.t Bladder not completely emptied before, catheter in-

serted at this point.

TABLE IV

Lack of correaion between carbon dioxide combining.Power and blood ketone levl

Blood carbon dioxidecombining power Blood ketones

oumes per cent mgm. per cent10 8013 7514 6416 150+17 86, 78, 3020 75, 2423 3727 7528 69, 1134 51, 60, 2544 1046 26

ma'rked fall in output with blood levels under 20to 30 mgm. per cent. (The exact levels varyfrom patient to patient.) In contrast, when ureaexcretion per hour is plotted against the bloodurea level, a straight line passing through the pointof origin is obtained. When the relationship be-tween blood and urine ketones is expressed as a

ratio: urinary ketones mgm. per hour the factblood ketones mgms. per cent

that the ketone bodies are'threshold substances isclearly demonstrated, as there is a striking changein the ratio at low and high blood concentrations.This, again, is in distinct contrast to thresholdsubstances, like urea, where the threshold is pro-portional to the blood concentration. Addis andDrury (10, 11) showed with urea "That undercertain special conditions the rate of urea excre-

TABLE V

Kton excretion in patient with oliguria and asotemia

Urine FuditkTime Blood ketones Urine ketones volume Blood sugar Cos C.P. Fluid intake

mgm. per cent mgm. per hour* cc. per hour mgm. per cent voumes per centt9:30 A.M. 1000 cc. N.S. i.v.

10:10 930 (N.P.N.-80 1610:10 to 11:25 ? 8 mgm. per cent)10:40 150+ 1000 cc. N.S. i.v.11:25 to 12:30 61.7 5511:45 1000 cc. N.S. i.v.12:15 P.M. 1000 cc. N.S. subcut.12:30 150+ 740 (chlorides-335 16 1000 cc. M/6 sodium12:30 to 2:30 455 150 mgm. per cent) lactate

1:33 1082:30 to 4:45 147 533:42 69 932 284;45 500 cc. blood transfusion5:55 39

* All urines gave a positive qualitative acetone test.

t CO, C.P.-carbon dioxide combining power.

t Just at death.

N.P.N.-non-protein nitrogen.N.S.-normal saline.i.v.-intravenously.subcut.-subcutaneously.M/6-sixth molar.

240


tion becomes directly proportional to the bloodurea concentration, so that in any one individual

urea in 1 hour's urine .the ratio: . sacntturea in 100 cc. of bloodwith only narrow limits of variation, over a widerange of blood urea concentrations." The factthat the percentage of acetone (and hence aceto-acetic acid) increased proportionally to beta-hy-droxybutyric acid at a very low blood level sug-gests that these substances are non-thresholdsubstances, and would account for the majorityof ketone bodies found in the urine under bloodlevels of 20 mgm. per cent. This is further evi-dence that beta-hydroxybutyric acid is a thresholdsubstance. It thus appears to belong to the groupof threshold substances, and behaves similarly tosuch electrolytes as chlorides (12), substanceswith "fixed" thresholds.

These results are of significance clinically. Thequalitative urinary acetone test, in the majorityof cases, roughly paralleled the quantitative totalurinary ketone output. The results in one patientwith uremia and oliguria suggest, however, thatsuch a gauge of blood Jevel, i.e. urinary acetone,can be used only in the presence of adequate renalfunction. This has been suggested by others (13,14), although Briggs (15) claims that it is im-possible to have ketone bodies in the blood andnot in the urine, due to diffusion directly from theblood to urine in the urinary bladder.

SUMMARY

Studies on 20 diabetics in mild and severeketosis have shown:

1. The low urinary output of ketone bodies(less than 100 mgm. per hour) with blood ketonelevels under 20 mgm. per cent. This is prob-ably due largely to the excretion of acetone whichappears to be a non-threshold substance.

2. A relationship between blood ketone levelsand urinary output:

a. Increased urinary output per hour withrising blood ketone levels.

b. A renal threshold for beta-hydroxybutyricacid over 20 mgm. per cent.

3. Impaired urinary excretion with renal fail-ure.

4. A lack of correlation between the blood

ketone level and the carbon dioxide combiningpower.

Appreciation is expressed by the authors to Dr. D. R.Drury and Dr. Eaton MacKay for criticism and help,and to Mr. Anthony Muff for technical assistance.

BIBLIOGRAPHY

1. Barnes, R. H., and Wick, A. N., A method for thedetermination of blood acetone bodies. J. Biol.Chem., 1939, 131, 413.

2. Van Slyke, D. D., The determination of B-hydroxy-butyric acid, acetoacetic acid, and acetone in urine.J. Biol. Chem., 1917, 32, 455.

3. Barnes, R. H., and Drury, D. R., Utilization of ketonebodies by the tissues in ketosis. Proc. Soc. Exper.Biol. and Med., 1937, 36, 350.

4. Peters, J. P., and Van Slyke, D. D., QuantitativeClinical Chemistry. Vol. I-Interpretations. Wil-liams and Wilkins, Baltimore, 1931.

5. Allen, F. M., Stillman, E., and Fitz, R., Total dietaryregulation in the treatment of diabetes. Mono-graphs of the Rockefeller Institute for MedicalResearch, New York, 1919.

6. Wilder, R. M., Intravenous injections of Beta-hy-droxybutyric and acetoacetic acids. J. Biol. Chem.,1917, 31, 59.

7. Briggs, A. P., and Shaffer, P. A., The excretion ofacetone from the lungs. J. Biol. Chem., 1921, 48,413.

8. Widmark, E. M. P., Studies in the acetone concen-tration in blood, urine, and alveolar air. I. Thepassage of acetone and aceto-acetic acid into theurine. Biochem. J., 1920, 14, 364.

9. Shipley, R. A., and Long, C. N. H., Studies on keto-genic activity of anterior pituitary; relation ofketonaemia to ketonuria in rat; method for assayof ketogenic activity; nature of ketogenic prin-ciple. Biochem. J., 1938, 32, 2242.

10. Addis, T., and Drury, D. R., The rate of urea ex-cretion. The effect of changes in blood ureaconcentration on the rate of urea excretion. 3.Biol. Chem., 1923, 55, 105.

11. Drury, D. R., The effect of very high blood ureaconcentrations on the rate of urea excretion. 3.Biol. Chem., 1923, 55, 113.

12. MacKay, E. M., and MacKay, L. L., Relation of theurine chloride rate to the plasma chloride con-centration before and after administration ofsodium chloride. Am. J. Physiol., 1936, 115, 455.

13. Richardson, R., Diabetic acidosis with negative re-action for diacetic acid in the urine. M. Clin.North America, 1932, 16, 257.

14. Labbe, M., and Boulin, R., Coma diabetique sansreaction de Gerhardt. Bull. et mem. Soc. Med. d.h6p. de Paris, 1933, 49, 313.

15. Briggs, A. P., The management of diabetes as con-trolled by tests of acetone in expired air. J. Lab.and Clin. Med., 1940, 25, 603.

241