An Improved Method of Diagnos- 
ticating Diabetes from a 
Drop of Blood. 
BY 
L. BEEMEE, M. D., 
ST. LOUIS. 
« 
REPRINTED FROM THE 
Ttfeto Yocfe Jf&etitcal journal 
for March 7, 1896. .Reprinted from the New York Medical Journal, 
for March 7, 1896. 
AN IMPROVED METHOD OF 
DIAGNOSTIC ATING DIABETES FROM 
A DROP OF BLOOD. 
By L. BREMER, M. D., 
ST. LOUIS. 
In two previous articles (Centralblatt f. d. med. Wis- 
sensch., 1894, No. 49, and Philadelphia Medical News, 
February 9, 1895) I have described a method enabling one 
to diagnosticate diabetes by the blood. The typical and 
practical feature of this process is that a drop taken from 
the tip of the finger in the usual manner by means of a 
needle permits of an unfailing diagnosis. I have, since 
publishing the article mentioned, continued my studies, and 
base the conclusions which I am going to give below on 
about fifty cases. These comprise both genuine diabetes 
as well as cases of glycosuria. The test mentioned above 
has been employed only by a few observers that I know of. 
Some have communicated with me, and the impression I 
have gathered from their letters is that its execution is a 
rather tedious and irksome one, even to the patient micros- 
copist. The difficulty of successfully repeating the experi- 
ment lies in the complicated process requisite to the prepa- 
ration of the staining fluid. This difficulty is enhanced by 
the fact that the two aniline dyes employed for the test— 
eosin and methylene blue—vary considerably according to 
their source. The acidity of the former and the alka- 
linity of the latter are met with in shifting proportions in 
the different articles furnished 'by different manufacturers. 
Again, the process of preparing the blood specimens by 
Copyright, 1896, by D. Appleton and Company. 2 
DIAGNOSTICATING DIABETES 
means of prolonged heating has proved objectionable to 
many who otherwise would resort to this test. This pro- 
cess consists in subjecting the dried blood specimen (cover- 
glass preparation) to a heat of 125° to 130° C. for two hours. 
This is time-consuming work requiring proper apparatus. 
The copper strip or copper plate ordinarily used is unre- 
liable. The danger of overheating the specimens is very 
considerable. In this event the test will be negative or 
doubtful. For these reasons I have tried to simplify the 
method, and now I am able to announce that I have suc- 
ceeded in devising a quick and practical process by means 
of which diabetes can be demonstrated almost as readily 
by a drop of blood obtained from the tip of the finger by 
the prick of a needle as it can be done by the urine tests 
ordinarily employed. Not only is my method nearly as 
quick as the chemic urinary test or polariscopic one, but it 
is, I believe, more reliable. Another improvement on the 
original method is that the microscope is rendered su- 
perfluous. Naked-eye inspection is sufficient to make the 
diagnosis. 
It is a well-known fact that by means of dieting and by 
the administration of certain drugs (antipyrine, for instance, 
calomel, and ammonium carbonate) the sugar can be made to 
temporarily greatly diminish or entirely disappear from the 
urine, even in cases of well-established and undoubted dia- 
betes. Fasting is a tolerably certain means of freeing the 
urine from sugar. But even in such cases, where the urine 
test is negative chemically and by the polariscope, my 
method reveals the diabetic, as I have demonstrated in sev- 
eral instances. The importance of such a test from a 
clinical point of view, as from that of examinations for 
life insurance, is obvious. I know from my own observa 
tion that diabetics are not infrequently accepted by life- 
insurance companies as good risks. The so-called danger 
line of specific gravity, T030, is delusive. There are dia- 
betics whose urine is, at times at least, of normal specific 
gravity. 
Diabetes is a much commoner disease than is gener- 
ally assumed by the profession and the laity. There can, 
furthermore, be no doubt as to the constant increase of 
diabetic patients, a fact that has been justly attributed to FROM A DROP OF BLOOD. 
3 
the increased strain which modern civilized man is sub- 
jected to. It is therefore of importance to have a means of 
diagnosticating incipient, or threatening, or doubtful dia 
betes. This, I maintain, can be done from a drop of blood, 
which, unlike urine, is always obtainable. 
The component parts of the reagent are eosin and 
methylene blue. Saturated watery solutions of these dyes 
are mixed in about equal proportions, so that a neutral, 
point is obtained; a precipitate forms which is insoluble 
in water but soluble in alcohol. This precipitate is washed 
and dried on a filter and reduced to a fine powder. To it 
are added eosin and methylene blue in small quantities. 
The amount of each varies according to the respective de- 
grees of acidity and alkalinity of the dyes. Owing to the 
variability and unreliability in this respect (there are at 
least four different kinds of eosin to my knowledge), the 
quantities have to be ascertained by practically and experi- 
mentally testing specimens of diabetic and non-diabetic 
blood. I will state that of the samples I had at my dis- 
posal, one twenty-fourth of eosin and one sixth of methy- 
lene blue was added to the neutral (dried) compound. This 
yields a powder of a reddish brown color. 
The manner of procedure to make the test is as follows : 
The drop of blood is procured by quickly pricking the 
fourth finger of the right hand of the person under exami- 
nation. The drop is spread either in the usual manner, 
being placed between two cover glasses, which are with- 
drawn from each other by means of the fingers or the for- 
ceps, taking care to obtain a film of equal thickness (which 
sometimes is hard to do) ; or the drop is placed near one 
of the edges of the square cover slip, held with the left 
hand, while with another cover slip, held with the right at 
an angle of about forty-five degrees, the drop is gently 
spread and then allowed to dry. It is now, for the sake of 
comparison, placed together with a cover-glass slip prepara- 
tion of non-diabetic blood in a wide-mouthed bottle or 
glass jar, such as is usually employed for portable galvanic 
batteries, containing equal parts of alcohol and ether, say 
about eight or ten grammes of each. This jar is then 
placed in hot water contained, say, in a tin cup, where the 
ether-alcohol is allowed to boil for four minutes. The 4 
DIAGNOSTICATING DIABETES 
boiling point is about 60° C. This is done for the purpose 
of fixing the haemoglobin in the red blood-corpuscles. The 
cover slips thus prepared are now transferred to the test 
fluid prepared by dissolving about 0'025 to 0-05 of the 
powder in about ten grammes of thirty-three per cent, of 
alcohol.* 
I must, however, call attention to the fact that this 
solution retains its characteristic staining qualities only 
for a few hours, and that for every examination a freshly 
made solution should be employed. In this fluid the cov- 
er slips remain for about four minutes. They are then 
washed in water, after which it will be found that the dia- 
betic or glycosuric blood film presents a sap or sometimes a 
bluish-green color, whereas the non-diabetic blood prepara- 
tion looks reddish-violet. 
The contrast is very striking, and is as plain as that 
which exists between diabetic and non-diabetic urine treated 
after the usual method, say with Nylander’s solution. 
This specific reaction I have obtained in about fifty 
cases of diabetic and glycosuric blood. (Herein are in- 
cluded, however, the cases examined with the method pre- 
viously published.) Usually the blood test was resorted to 
first and was followed up, viz., verified, by the urine test. 
Even in such cases where there is only a trace of sugar 
demonstrable in the urine—for instance, where a slight 
black precipitate is noticeable with Nylander’s solution 
only after settling on standing for some time—the test is 
positive. Lastly, there is a class of persons who are not 
only predisposed to diabetes, but dwell actually on the 
borderland of this disease, crossing and recrossing the bor- 
der line. They do not habitually show sugar in the urine. 
These individuals I call “ sugar liners,” in analogy with 
the term “border liners.” This “diabetic border liner” or 
suspect can also be told by my method. Nothing, however, 
is easier than to pick the diabetic out of any number of per- 
sons, well or sick with any other sort of ailment, by means 
of my method. 
* Alcohol of higher grade gives rise to the formation of a great 
number of small air bubbles formed between the erythrocytes and the 
plane of the cover glass, and in the protoplasm of the blood-corpuscles, 
thus interfering with the study of the corpuscles under the microscope. FROM A DROP BLOOD. 
5 
The peculiar reaction described consists in the selective 
affinity of a certain substance occurring in red blood-cor- 
puscles of diabetic patients for another substance contained 
in the color reagent; possibly a specific principle or sub- 
stance in the diabetic red blood corpuscles forms a new 
chemical compound with one of the staining principles 
contained in the eosin-methylene- blue solution. At all 
events, the red blood-corpuscles of diabetic or glycosuric 
blood are stained green, whereas those of non-diabetic 
blood present a purple or madder color. There is no dif- 
ferential staining of the blood plasma, except in so far as 
the exuded fluid from the red blood corpuscles resulting 
from the drying process is concerned. The test proves 
that there must be a substance other than grape sugar on 
which the characteristic stain depends ; for a cover glass 
smeared with non-diabetic, say healthy blood, when pre- 
pared in the usual manner and treated with a solution of 
grape sugar, does not yield the diabetes reaction on the 
application of the reagent, nor will the blood of an animal, 
say a rabbit, that has received two grammes of grape sugar 
subcutaneously yield the specific reaction within one or two 
hours after the injection. But the specific reaction is ob- 
tained by floating a cover-glass slip with a film of non-dia- 
betic blood on diabetic urine for about ten or fifteen minutes. 
Here the stain is, too, the characteristic green color. In 
this case, then, a substance contained in the red blood- 
corpuscles, perhaps the haemoglobin, absorbs and incorpo- 
rates, possibly combines with, the substance that occurs in 
the urine of diabetes, determining the same chemic reac- 
tion as we see in diabetic blood. By treating, then, nor- 
mal blood with diabetic urine the former can be rendered 
diabetic so far as the typical color reaction is concerned. 
Whether this substance is directly toxic or only is instru- 
mental in interfering with oxidation is a question. It is 
reasonable to infer that the latter is a well-grounded con- 
clusion, being a foreign substance lodged in the oxygen- 
carriers. 
I may state further that I have experimented with all 
kinds of diabetes and glycosuria, so far as the probable or 
supposed origin was concerned. The results have been 
uniform whether the disease was one of the hepatogenous, 6 
DIAGNOSTICATING - DIABETES 
haimatogenous, neurogenous and psychogenous, traumato- 
genous, pancreatic, or toxic varieties. In all of them the 
reaction will infallibly be obtained. 
I would in particular call attention to artificial or experi- 
mental diabetes produced by the administration of phloro- 
glucin to an animal. Of late I have experimented with 
rats. The phloroglucin rats show the typical reaction, both 
as to blood and urine. In rats, however, treated with 
phlorrhizin the specific blood reaction is absent, although 
the urine shows a sugar reaction with Nylander’s solution. 
The action, therefore, of phloroglucin on the organism is 
totally different from that of phlorrhizin. (I did not repeat 
the experiment of floating specimens of non diabetic blood 
on the urine of phlorrhizin animals.) Whereas the former 
seems to give rise to all the metabolic abnormities of the 
blood met with in genuine diabetes, including the white, 
unstainable, necrotic (?) corpuscles described in my former 
articles, there is an absence of that peculiar substance in 
the erythrocytes which gives the characteristic color reac- 
tion with the test dyes. This substance is not grape sugar, 
as remarked before. 
Another point of interest is that the lower vertebrates 
are susceptible of toxic diabetes, as has been demonstrated 
by other observers. At all events the chicken is. A 
chicken cock was given phloroglucin, and the diabetes reac- 
tion of the erythrocytes could be demonstrated after three 
days. Unfortunately, I had no frogs at my disposal, and 
had therefore no opportunity of ascertaining the positive 
or negative reaction on a still lower rung of the scale of 
evolution. 
Again, it is perhaps not unimportant to state that the 
blood of the chicken embryo, taken on the fourth and fifth 
day of incubation, shows the same color reaction of the 
discoplasm as diabetes blood. I refrain from drawing con- 
clusions and advancing theories. 
Among the patients whose blood 1 have examined, the 
diagnosis of diabetes (and glycosuria) was generally made 
by the blood test first, then verified by chemical and polari- 
scopic examinations. Sometimes, when the latter did not 
reveal any sugar in the urine, owing to the small percent- 
age, the blood would tell the tale, larger and more appre- 7 
ciable quantities being found on successive trials. This 
would, in a measure explain the observation that the degree 
of gravity of diabetes is often independent of the amount 
of sugar found in the urine. The sugar may disappear 
from the urine, but the toxic state of the blood persists. 
I have not succeeded yet in diagnosticating diabetes 
from glycosuria. Possibly this can be done by perfect- 
ing the blood-testing methods. I have no doubt that 
there are other, perhaps simpler, aniline stains than those 
which I have described. The Ehrlich-Biondy stain, for 
instance, imparts a light yellow color to diabetic ery- 
throcytes in contradistinction to the brownish-yellow of 
the non-diabetic blood. With this reagent, too, a naked- 
eye diagnosis could be made with a little practice, but the 
contrast is not so striking as that obtained with the re- 
agent described above. 
There is one point which I should like to call particu- 
lar attention to in order to prevent misinterpretation. 
Non-diabetic blood when spiead in an uneven manner, so 
that ridges or excessively thick layers are formed, may also 
show a greenish tint. Again, the blood film that is not 
sufficiently hardened either by the ether-alcohol or by heat 
may give doubtful results. Lastly, the dichroism of the 
erythrocytes must be taken into consideration—i. e., the 
fact that blood spread on glass appears red on refiected 
and greenish on transmitted light. But if these sources of 
possible error are avoided the test is absolute. 
Chloral or chloralamid glycosuria is not comprised in 
my experiments, nor is experimental curare diabetes. 
The method described has the disadvantage that minute 
crystals, which, during the staining process, are precipitated 
on the specimen, mar the latter somewhat for microscopic 
examinations. It seems that a status nascendi for the 
newly forming compound of eosin and methylene blue is 
-indispensable to the reaction. Solutions kept in ordinary 
alcohol lose the quality of the differential stain. Hence 
the injunction to prepare a fresh solution for every test to 
be made. The older the test solution is the more uniform 
—i. e., the less differentiated—is the color of blood speci- 
mens of diabetic and non-diabetic blood. Both assume a 
bluish tint. In order to obtain the finer histologic features 
FROM A DROP OF BLOOD. 8 
DIAGNOSTICATING DIABETES 
of the blood, diabetic or non-diabetic, especially the differ- 
ential staining of the nuclei and the granulations of the 
white blood-corpuscles, ten minutes’ staining is requisite. 
The chief results of my investigations of the blood in 
diabetes, both clinical and experimental, briefly stated, 
are : 
1. The diagnosis of diabetes from a drop of blood can 
be made with as great a certainty, perhaps with a greater 
one, and almost as quickly, as by the urine tests. 
2. There is a substance, at present unknown, which oc- 
curs in the diabetic red blood-corpuscles, foreign to these 
bodies in the physiological state, which causes the specific 
reaction with the eosin-metliylene-blue compound de- 
scribed. In no other condition of the blood is this sub- 
stance (or combination) met with than in diabetes and gly- 
cosuria. The only exception 1 have met with so far is in 
embryonic chicken blood. 
3. It is not the presence or excess of sugar in the blood 
which causes the clinical symptoms of diabetes, but it is 
with greater probability the foreign element alluded to, 
which is probably combined with the haemoglobin, in con- 
junction with the white (necrotic) masses resulting from the 
decay of the corpuscular elements of diabetic blood, that 
makes the clinical symptoms and is possibly the anatomo- 
pathological substratum of the disease. 
Postscript.—The results of the blood-tests detailed above had 
been so uniform that I was firmly of the opinion that in every case 
where sugar could be demonstrated in the urine the characteristic 
blood-reaction could be obtained. In one case, however, the test has 
been negative, although it is clearly one of diabetes, showing 6'5 per 
cent, of sugar. It is the case of a boy, sixteen years old, who, at the 
age of fourteen, became diabetic in consequence, as he alleged, of an 
electric shock. I shall describe this case in detail. A cover-glass 
preparation of normal blood floated on the urine of this patient did not 
show the characteristic green color. Since the blood-reaction was 
normal, I take it to be analogous to phlorrhizin or essential kidney dia- 
betes. There are, then, at least two distinct kinds of diabetes that can 
be diagnosticated by the blood-test—haematogenous (broadly stated) 
and renal diabetes. The recent experiments of an Italian observer 
with oxygen on diabetics would practically bear out the conclusions 
arrived at by me—that, with rare exceptions, diabetes is a disease of 
the red blood-corpuscles. It seems reasonable to infer that an abun- 
dant supply of oxygen—for instance, inhalation of oxygen—will benefit FROM A DROP OF BLOOD. 
9 
those patients in whom the oxygen-carriers are hampered in their 
function. 
Description of Plates A and B. 
The plates represent cover-glass preparations, of A (with 
madder-colored erythrocytes), normal, B (with green-colored 
erythrocytes), diabetic blood. The specimens were heated 
for two hours to 125° to 130° 0. (The ether-alcohol process 
yields the same pictures.) They were stained in the same 
eosin-methylene-blue solution according to directions given 
in the preceding article. 
a, leucocyte (epsilon cells—Ehrlich); Z>, eosinophilous cells 
(alpha cells—Ehrlich); c, smallest lymphocytes, showing a 
delicate seam of protoplasm; d, medium-sized lymphocytes, 
with basophilous granulations; e, large lymphocytes; /, plate- 
let patches. The platelets (smallest blue bodies) in A show 
small white globules: in B these are larger and of irregular 
shapes, g, red blood-corpuscles. Those marked with letters 
on Plate A have small white dots in the central portion 
which I have called stigmata of the erythrocytes (see Arch. f. 
miJcrosJc. Anatomie, vol. xxxv, p. 441). It will be seen that 
some of the red blood-corpuscles have distinct dots, colored 
the same as the protoplasmic bodies of the erythrocytes. 1 
consider them the atrophied or senile nuclei of the red blood- 
corpuscles. The erythrocytes show varying degrees of tinc- 
torial power in the normal as well as in diabetic blood. THE 
New York Medical Journal. 
A WEEKLY REVIEW OF MEDICINE. 
EDITED by 
FRANK P. FOSTER, M. D. 
THE PHYSICIAN who would keep abreast with the advances in 
medical science must read a live weekly medical journal, in which 
scientific facts are presented in a clear manner ; one for which the 
articles are written by men of learning, and by those who are good and 
accurate observers ; a journal that is stripped of every feature irrelevant 
to medical science, and gives evidence of being carefully and conscien- 
tiously edited ; one that bears upon every page the stamp of desire to 
elevate the standard of the profession of medicine. Such a journal ful- 
fills its mission—that of educator—to the highest degree, for not only 
does it inform its readers of all that is new in theory and practice, but, 
by means of its correct editing, instructs them in the very important yet 
much-neglected art of expressing their thoughts and ideas in a clear and 
correct manner. Too much stress can not be laid upon this feature, so 
utterly ignored by the “ average ” medical periodical. 
Without making invidious comparisons, it can be truthfully stated 
that no medical journal in this country occupies the place, in these par- 
ticulars, that is held by The New York Medical Journal. No 
other journal is edited with the care that is bestowed on this ; none 
contains articles of such high scientific value, coming as they do from 
the pens of the brightest and most learned medical men of America. A 
glance at the list of contributors to any volume, or an examination of 
any issue of the Journal, will attest the truth of these statements. It 
is a journal for the masses of the profession, for the country as well as 
for the city practitioner ; it covers the entire range of medicine and sur- 
gery. A very important feature of the Journal is the number and char- 
acter of its illustrations, which are unequaled by those of any other journal 
in the world. They appear in frequent issues, whenever called for by the 
article which they accompany, and no expense is spared to make them of 
superior excellence. 
Subscription price, $5.00 per annum. Volumes begin in 
January and July. 
PUBLISHED BY 
D. APPLETON & CO., 72 Fifth Avenue, New York.