Diabetes mellitus in pregnancy

examiuation showed there was no &rough the vagina if possible and an incision over the preswting I)xrt The operation. was cxecedingly easy requiring a ligature. After the removal of the tumor, small fibroids on the fnndns the uperation was necessary. 625 EUI

pedi& to the tumor. E decided ti remove it By making to do a hyaterc~tomg ii ucecssa~ry. ujith a blunt dissector, ti1C tIllnor bYi% cancleated. and. almost bl.oodleas, only une bleeding point it was

easy

uterus

wars

to determine otherwise

that normal,

excepting the two and no further

STILEET.

HE carbohydrate metabolism ol: the normal pregnant woman maabe said to be in a more labile condition than in the nullpregnant. This is evidenced by the comparatively frequent sppearanee of traces i;f glucose or lactose in the urine, and also by the greater tendency to acidosis in the pregnant state. It seems reasonable to ascribe this mstabilEty to the c.hanges in structure and probably also in function -Lvhich pregnancy is known to induce in certain organs of internal secretion. The thyroid, the anterior lobe of the pituitary, as well as .the ehromaffin system hypertroph.y in pregnancy, and the increased activity of these sugar-mobilizing organs might be expected to affect the glycolytie function of the pancreas. There is also apparently a reciprocal influence between the gonads and the pancreas, for in diabetes mellit.us disturbances of sexual function such as amenorrhea, sterility, and impotence are common. Changes in the islands of Langerhans in pregnancy have been described, and similar changes have heen found in the pancreas in eastrated animak. The pancreatic activity is no doubt mod.ified in normal pregnancy, yet permanent pancreatic damage is very rare. In fact i.t is doubtful whether diabetes mellitus ever develops in pregnancy uncomplicated by toxemia. Glycosuria occurring durin g the course of pregnancy may indicate the existence of one of three separate clinical entities, namely, (l), renal glycosnria; (2), a type of diabetes peculiar to pregnancy; and (31, true diabetes mellitus as evidenced by the classical symptoms and clinical. course. Lactosuria, which has no significance, occasionally occurs towa. i,he end of pregnancy. It is, therefore, necessary to apply the diPferential phenylhydrazine and fermentation tests to a reducing substance in the urine of the pregnant woman. .-*Read at a Inc@ting of the Slosne

TIoegitnl

Alunmi

Society, Octobw 26, 1923.

WIENER

:

DIABETES

MELLI’I’US

IN

PREGNANCY

711

1. Renal glycos,uriu.-The not infrequent appe#arance of traces of glucose in the urine in pregnancy, with a normal blood sugar level, may be attributed to a lowering of the renal threshold. Such a typical renal glycosuria has been noted to occur in nonprcgnant individuals in the presence of focal infections, and to, have disappeared when these foci have been eliminated. Possibly end-products of the fetal metabolism may in like manner affect the renal threshold o-f the maternal kidney. The proven changes in the concentration of the inorganic constituents of the maternal blood during pregnancy, particularly calcium, also might be a, factor in the alteration of the renal threshold. This low renal threshold for glucose is so constant a feature of normal pregnancy, that it has been successfully utilized as a diagnostic test for pregnancy, by means of the administration of sugar by mouth, or the injection of minimal quantities of phloridzin or adrenalin. The amount of sugar excreted in cases of renal glycosuria is usually from 0.2 to 0.3 per cent, amounting to 3 to 5 grams in twenty-four Increased carbohydrate intake has little or no effect. A hours. glucose tolerance test in these patients yields a normal blood sugar curve; in fact frequently the blood sugar level does not even rise to the usual level of the nonpregnant. The carbohydrate allowance of these patients should not be reduced. The glycosuria disappears promptly after delivery. 2. Transient Di.abete,s Mellitws of Pregnancy.-The second group of glycosurias in pregnancy is represented by cases which develop a hyperglycemia and a high concentration of glucose in t,he urine, without clinical symptoms of diabetes mellitus. Two such cases have been observed among the admissions to th.e Sloane Hospital during the past two years. In both cases there were definite symptoms of a mild toxemia of pregnancy. The first of these, a colored woman, gravida iv, (age 22, Ht. 65 inches, Wt. 227) had a history of three normal pregnancies and normal deliveries. She was admit,ted to the hospital twelve days before delivery because of the presence of edema of the legs and abdominal wall and a blood pressure of 132/70. A trace of albumin kvas found in the urine on the second day after admission. The fasting blood sugar was then 91 mg. per cent with a trace of sugar in the urine at t;he time. The 24hour collection of the same day, 1020 cc., showed a concentration of 2 per cent of glucose, a total excretion of 20.4 gm. The diet, which contabned approximately :!OO gm. carbohydrate, was not modified, and two days later the fasting blood sugar was 133 mg. with 2,.3 per cent in the 24-110~~ collection of 1500 c.c., or 34.5 gm. On the following day the 24.hour urine amounted to only 540 LX., yet contained 4.4 per cent glucose, or 37 gm. On the next day, two and one-half hours after an intake of 36 gm. CH, the blood sugar was 250 mg.: and the urine voided at the time contained 9.2 per cent glucose. The CH intake was then reduced to 100 gm. per day. On the following day the patient was delivered of a stillborn child, (breech presentation). The autopsy diagnosis was cerebral edema and congestion. Two days postpartum the urine contained 0.75 per cent glucose, 4.9 gm. in twenty-four

hours. The fasting blood ;Kir cent glucose in the zeretion

ill

was

mg.,

with

sugar on the fifth day postpartum was 130 mg. with 3.7 urine at the time. Nine days postpartum the 24.hour On the eleventh day postpartum the Basting blood sugar was cent glucose in the urine at the time.

8 gm.

0.2 per

This patient was discharged in excellent condition 16 days postpartum. Throughopt there had been no polyuria, polydypsia, or other clinical signs of diabetes mellitus. Ketone bodies were never found in the urine. Three months later, during which inter& she had been on an unrestricted diet, the blood sugar taken one and one-half hours after an intake of 100 gm. of CK was 133 mg., with no sugar pre%?Ilt ill the urine. The second case of this type, a colored woman of thirty-eight, (Kt. 64 inches, Wt. 208) parn v, gravida ix, admitted September al,> 1923, had had two preyious severe attacks of t.oxemia of pregnancy. The last attack, which was aeeompanied by persistent pernicious vomiting and jaundice, necessitated a therapeutic abortion at three and one-half months in September, 1922. Previous to this operation, she had received two intravenous 10 per eent glueose infusions two da.ys apart. The rirst infusion contained 110 gm. of glucose, the second 80 gm. Following the operation 0.5 per cent glucose appeared in the urine, which again became sugar-free .five days postoperative.

I ,~

BLOOD

PLJ u3: ia

24

I

HOUR

URINE

I

3 Fz iG.” -166 .. . ... . .. . .. . .. 138 e. . 123 0..

iii 100 111 101 101 -- 146

-

6 d f SIG.”

27.8 .. . .. . . .. . .. . .* 28.6 .. . 23.6 .. . .. . .. . .*. . .. 18. .. . .. .

j!

3:; ..

i.i

1::: ..“f

255 . ~. ..I .~. e.. ... ,.. ... 235 .~. :::

i / I i

900 700 1050 550 1050 2050 700 1000 1400 1000 1250 1050

She again became pregnant in March, 1923, and was September 21 in the seventh month, because of glycosuria Her fasting blood sugar on admission of mild toxemia. The Whour urine contained 4 per cent glucose, 36 gm. Her COH time she had been on an unrestricted diet. to approximately 60 gm., and the urinary sugar decreased zent on discharge twelve days later. The 24.hour volume blo,od sugar IeveI dropped 2000 cc. The fasting nom placed on a diet containing 100 gm. COH, and on two

4. 1.43 1.74 1.17 0.1 Tr. 1. 0.38 0.1 h. tr. 11. tr. 1.1 . .. . . .. .

36. 10. 18.3 6.5 1. *.. 7. 3.8 1.4 ... I.. 11,5 I.. e. .

R E 2 R ____ 0 0 f

ff 0 0 0 0 0 0 0 0 s. ..

admitted to the hospital and subjective symptoms was 166 mg. (Table I.). in 900 GX. Up to this intake wa,s now reduced progressively to 0.2 per varied between 900 and to 121 mg. She was subsequent visits during

WIENER

:

DIAUETES

MELLITUS

IN

PREGNANCY

715

October the blood sugar two hours after breakfast of 40 gm. COH was 100 and 111 mg. The urine contained a trace of sugar at the time, as did the 24 hour specimens. She was readmitted October 25, because hypertension had developed and albumin was found in the urine. The fasting blood sugar was then 100 mg. with no sugar in the urine at the time. She was again. discharged after two weeks of treatment for the toxemia, and during this period there was no glyeosuria with a diet containing 200 gm. COH. She was readmitted November 15, and delivered November 17 of a living child. On the third day postpartum the fasting blood sugar was 146 mg. with a trace of glucose in the urine at the time. Traces of glucose were found

in the urine (0.1 to 0.2 per cent) up to the day before discharge, when it cleared up entirely.

November 30,

These cases resemble those described in the literature in which glyeosuria of marked degree has occured during a pregnancy to disappear again postpartum. In some the sugar in the urine was again found to be present in a succeeding pregnancy, while in others it did not recur. The accompanying toxemia in the two cases detailed above was of mild degree, but should be considered as possibly etiologically connected with the glycosuria. The fact that both patients were obese women should also be noted. Apparantly this form of glycosuria is caused by a transient functional insufficiency of the islet tissue of the pancreas. Functional recovery appears to be complete after delivery. 3. Pregnancy in Cases of Did&es MeWus.-Von No’orden’s statistics of 427 diabetic women in the childbearing period have shown that pregnancy occurs in only 5 per cent. Although in the more severe cases conception is less frequent than in the milder form, the incidence of sterility does not run parallel to the severity of the disease. My own experience is confirmatory of the above, for of the twenty diabetic women in the childbearing period followed at the Vanderbilt clinic during the past six years, only one, and that one a severe case, became pregnant. It is presumed, although without definite pathologic evidence, that diabetes may cause sufficient damage to the ovaries to Statistics of 58 cases reported in 1909 bring on complete st,erility. showed a 30 per cent mortality from diabetic coma either during labor or immediately postpartum, and a 21 per cent mortality within two and one-half years following labor. AL number of cases of diabetes have been reported in which a slight improvement was observed durin tile condition, but the ing pregnancy, or at least no progression majority of observers have noted the distinctly detrimental effect of a pregnancy. The intrauterine mortality for the child has been stated to be 50 per cent, and in codrast to the improvement in the maternal condition after death of the fetus in the various toxemias of pregnancy, in the diabetic death of the fetus was frequently followed soon after by the death of the mother in coma. A certain number of premature a.nd weakly infants some with hydrocephalus and congenital diabetes were born of diabetic mothers. Extreme hydramnios has fre-

II! 20 25' 30 7

‘( (‘ Cb I6 May

____..____ *I?ei

20 25 25 28 22

25

2‘2 30 30 17 21 15 50 14 20 25 25 25 25 25

15

20 20 15 15 15 15 15 15

20

-.__-. 100 C.C.

4.

Apr.

" I6 ‘I

IL ‘( Mar, ‘I

8

12l 15' 17 24; 26 2I 9 13 1 23 51 91 16i 19, 2X,

i

'-

6‘ '8 “ 6‘ CL Feb. II

28' 8/ 13, 20' 27' 6' 201 27;

‘( Sep. Nov. (‘ ‘( Dec. “ ‘I 1923

..--.~--.:..

~C.Q.H. CM.

,-

Jan.

21~

1922

I_------____

Aug.

_..__

_I.-

1540 IF90

1420 I(338 1528

120 150 110

28 45

1560

140 150

140

150

109

75 75 68 40 150 90

955 995 995 914 6-14 1602 1150 .I205 1590

49 Go

54

41 50 69 59 45 50 43 70 36 43

48

1 GO 51.

:

'

; :

41

IWO 1860

1860

: 41 41 41 41

’ 59

2040 2040 1860

39 47

55

C.O.H.

'lYJl?AL iAVAIL' ARm

-

1710

~.

75

i

2012 2000

1400

lC~LO~~~s i

.

180

180

200 200 150 200 200 180 180

120

PAT GM.

_-

50 60

50

45 45 50 50 54 50 48 35 42 40 50 50 50 50 50

33 30 75 45 45 45 45 45

60

PRQTEIN GM. - --

DIET

___--_,_-

IIOUR

49 59 30 33 13

83

52 98 73 65 55 85 46 83 106 48 56 84 67 101 53

17 25 17 23 23 20 42 46

38

GM.

GMJCOSE

24

-.---

(

1

/

/

j

;

I

i

i

)

! ~ I

\

:

,

/

iI

^-

0 f, t 0 t' t

t + +,

0,

0 0

0

.t + t tt 4-t 0 tt

trace trace

ttt

tt -it

traec

tract

:'

0

I )IACETIC

URINE

'

.. .. .. ..

29

11 45 :14 6 10 38 3 13 70 5 1 30 23 47 . .

^i. 5

:: . . .. . .

*.

NEGATIVI

I'_--

1

_G

CLLH. I

BAl

.

30 18 .15

2;

. s

.. 1

..

..

..

30 .30 A0

20 30

. .

.. .~ .. .. .. .. .. .. ..

.1 ..

.. ..

. .

.

..

.

.

._ ~

,

.

. ~

..

.. .*

!

.

.

.

.

.

1

1:: 27s 238

. ..

. ..

225 228

215

204

iii ... . .. 249 215 . .. .. *

! Z-L4 *-.

i

.

.. . . ..

.,.

.

.

! 210

.

.,.

.

.

.

I 3G4

. ... . ... . ... . ~.~ ~ ..1

. ...

49.3 . .** ~... . "~I . ... ~ .. .. . ... 50. . . . ~ ~~. 50.0 . ... . ... . ...

. ... .. . .. . . ... . ... . ... . ... . . . ...

0% 'OL. "7( -~-

LASMA

.

.

.

.

.

.

.

.

.

.

.

.

‘

.

~

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. .~

-t1.3

. .. .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.~.

.

.,.

..I

.

..”

.

.,.

.

.~.

I..

.

11.4

.

.

.

54.6 5G.2

57. 58.0 y 57.5

/

,

52.5

52.5 52.4

51.0 51.0 51.6

50.0

51.1 51.1 50.4 51.6 51.5 50.0 51.1 50.0

62.0

53.0

53.1

52.2 52.4 53.2

ciz.5

period

I

QY

Zletin 1Ou.b.i. / d. started. No iletin this

I coryza ( I 1 Coryza i

I

i

I

,

'

I8 k i p p e d

Nausea

~~enetruating ' ;A.iVL

30 30 40 40 40 40 40

40

27 28

28 28

II 15 20 29 11 16 20

25

“ ‘( ‘( (I July “ <‘

“

“ 27 Aug. 22 Xept. 14

“ 26 act. 5

“ 12 24 II i9 28 Nov. 2 26 “ 9 18 ‘( 16 30 dr 23 30 I' 28 16 Dec. 7 16 ____ *per 100 C.C.

54 43 54

33 30 30

“ 28 June 1 ‘I 6

51 50 36 59 57 60 42 42

60 55

50 55

54

54 54 54 54 54 54 54

50 50

-ixfw 14 ' 25 I "- 18 25

162 i50 205 223 142 150 102 - 82

100 90

80 120

160

150 1so 180 160 160 160 l60

200 170 150

12.0 100

Ml6

195G 1956 1996 1816 1816 1816 1816

2148 1866 1956

1380

1754 1662 2088 2315 lGl0 1710 1150 970

1252 1154

A dmitted 1028 1412

~--_______

44

53 10 17 26 39 19 54

41 52 26

10 80

0

0 0 0 t t 0 t

4-t t t

0 t

4oi 40

30 33

6o 55 42 39

46 52 63 65

~ ' ~

24 40

26 47

53 57~

62 61 : 0

0 3I +7-J tt t trace trace 0

tt

o Sloane HOSE &I in lal 30 r. 0 El I "1: 0

71

61 61 71 71 71 / 71 j 71 i

64 i 61 / 61 ~

1

c4 I

/ 54

3 .. .. .. ..

i4

..

.. ..

See T: .. ..

..

.. .. . . . . ..

.. .. ..

i;i

21 26 7 10

..

29 17

36 14

tble III.' 24 48

27

8 51 54 45 32 52 17

23 9 35

44 ..

.. .

10.8 . . .. . . .. 10.3 10.0 . . .. . . .. . . ..

. . ..

. ... . ... . . ..

. .. . I . .. . I

. ... . . .. . . .. . . .. . . .. . . .. . . ..

. ... . ... . ...

... ... ... 213 ... .. . 235 250 . .. .. .

. ... . ...

. .. ...

. . .. . ... 52.0 . . .. . . .. . ... . ... . ...

.. . ... 50 50 50 50 60 60 GO 60

2F2 .. . .. . ... ... 300

. . .. . . ..

. .. .. 50 35

. . .. a. . . 10.7 . . .. . . .. . ... . ... . ...

. . .. . ...

Discharged A plst 1: .* 10.5 40 305 . . .. 50 .. . . . .. . . ..

60

40 50 50 50 50 50 50

30 40 40

30 ..

59.2 OS,2 60.5 60.6 62.5

64.5

62.5 63.8

61.9 61.9

61.7

59.0 59.6

59.0 59.0

I

of B.P.

edema

Cnsulin in on0 injection

Cn country two weeks

Albumin faint trace

To

Fsace of Albumin

normal

legs,.

Edema

Yo insulin for 24 hrs.

q~mxtly bee11 the indic2ition for premature majority opinion in the past has recommended pregnancy at an early period iii its course.

iuduetion of Iabur. ‘I’iis interruption of the

In the foll~owing case the patient became pregnant two and three-quarter years after the onset of diabetes. There was no familial history of diabetes, and the patient’s own history up ta the time of her first pregnancy in 1919 was irrelevant. A report from the Lying-in Hospital states that she then suffered from a severe toxemia, with marked edema, hypertension, large amount of albumin: hpaline and She improved under treatmen@ hut ~wmular casts, but no #sugar in the urine. ct ieft against advice and was confined at. home a day later being delivered of a stillborn child @@ember 7, 1919. Convalescence was uneventful, and she felt perfectly well until the first symptoms of diabetes appeared five, months later, in Her weight at that time was 132. pounds. Her treatment nias February, 1920. directed by Dr. Geyelin from the fall of 1920 to July, 192,Z. When she first came to the Vanderbilt Clinic in August, 1922, aged twenty-four, she weighed 111 pounds, {Ht. 62 inches.) and had 40 gm. of sugar in 2400 EC. of urine, on a diet of C. 20, P. 60, F. 120. (Table II.) Her blood sugar was 364 mg. per 100 C.C. During the following three months she ga.ined four pounds and the urinary sugar decreased on a diet of 6. 15, P. 45, F. 200. She became pregnant in November, and the sugar excretion soon after rose to 50 gm. although she maintained the same diet. During February and early Narch the sugar output increased up to 100 sm., with large amounts of acetone: and her weight fell to 108 pounds. Clinical symptoms of acidosis did not appear, and the Go, combining power of the plasma never fell below 44 volume per cent. Insulin injections, (Iletin, Lilly, H-120), were commenced on l/larch lS, and she has received two or more injections daily, with a.n occasional omission, ever since. From 10 units twice daily the dose was gradually increased to 25 units twice a day on June 16. The sugar excretion varied between 10 and 50 gm. in twenty-four hours during this period. On July +a on a diet of C. 40, P. 54, F. 160, there Qwas 1.66 per cent of sugar in 2460 C.C. of urine, 44 gm. The body weight was hen 142 pounds, blood sugar 250 mg. per 100 C.C. The Iletin was increased to 30 units twice daily. The patient was admitted to the Bloanc Hospital in labor at 6 a. X. of July 37. (Table III.) The fetal heart could not be heard, and the resident physician, Dr. B’indley, stated that hydramnios was probably present. At 10 A. U. the blood sugar was 312, CO, 38. The cervix was then two fingers dilated. Folloting an attack of vomiting, with drowsiness and moderate hyperpnea, 300 c.e. of 10 per cent g1nco.x At 2 P. $5. the solution with 45 units of Iletin were given intra,venouslg at noon. Wood sugar was 245, the Co, had dropped to 28. At 4.30 she was, given 20 units of Iletin subcutaneously and milk with glucose by mouth. At 6 :30 P. 3111 she was delivered breech presentation, of a stillborn premature female infant, which was 57 cm. long and weighed 19’00 gm. The cord was not pulsating, and there’ n-as some indication that there had been premature separation of the placenta. At 8:30 F. MI. 30 units of Iletin were given subcutaneously. At 10 P. x. the blood sugar had in20 unit doses of Iletin were creased to 465, and the CO, was still 28. Additional given at 12:30 A. X1. and G:30 A. nf. on the 28th. At 10 si. w’. the blood sugar rran 260, and the CO, bad rie,en to 48. It was palpably evident that the patient was on -he verge of coma during the last few hours of labor and the first twelve hours :)ostpartum, and she would undoubtedly have ~uccumbled to it if insulin had not :Jeen available. She remained in the hospital for 12 days postpartum with the ava.ilable COH in the diet, (COH plus 58 per cent of the protein) between 36 and 8i gm. and 30

29 30 31 1 Aag. 2 Aug. 3 Aug. 4 Aug. 5 Aug. 6 Aug. 7 Aug. 8 Aug. 9 Aug. 10

28

July

July July July Aug.

27

P. M. P.M.

P.M.

Jnly

10:

Noon

12: 2: 4:30 6:

P.M.

A.M.

27

10:

July

39

33 48 30 42 3s 4s 50 50 45 52 53 53 53

37 25 42 30 36 35 35 35 50 4s 58 50 50

7 7 7 21

72

24 24 24 102

. . . . . .

DIET

-

65 60 64 60 65 90 90 90 90 80 90 54 54

52

8 8 8 24

.. .. ..

865 832 860 860 881 1146 1150 1150 1190 1120 1254 1168 1186

912

708

i& 196 196

ii0

56 62 60 54 58 64 64 64 76 78 89 81 81

79

28 28 114

as “

7 5 5 4.5 0 0 0 0 Trace 0 14.9 6.1 0

11

1.2% . .. . 1.8% . .. . . *. . . .. . 24 gm . . .

.

F.OT. 0

0 0 0 0 0 0 0 Trace F. T. T T. F. T.

0

tt . *. +tt . .. .. . . .. tt

URINE

f.... .. . .. .. . . . .. . .. . . .. . . . .. . . . . .. ii .. . .. . . . *. . . . .. .. . .. . . . ..

.. . ..

-. . . .. .. . .. . . . .. .. . .. ..*.. . . . .. -90 . . .. .

C.O.H. BALANCE

TABLE

49 57 55 50 58 64 64 F4 76 78 74 75 81

GS

.. .. .. .. .. ..

III

60 GO 60 60 60 GO 60 60 40 40 40 40 40

70

.. . .. .. .. .. 85 .. 20 20 30 30 bid 30 bid 30 bid 30 bid 30 bid 30 ‘bid 30 bid 30 bid 20 bid 20 bid 20 bid 20 bid 20 bid

INSULIN

..

166.

53.

..

334.

415

..

47.5

is.

ii. ..

38.

235.

260

ii; . . . .. 465

312.

10.5

10.5

9.G

..

..

I

post

after

day

14 hours after jection

5 hours

3 hours

labor In labor In labor Postpartum Total for

I.In

In labor

last

jnsulin

breakfast

insulin

in-

wits of lletin given twice daily. Four days before discharge the urine was sugarfree, but on an intake of only 880 oalories, (C. 25, P. 40, I!‘. 60.) Two days before discharge with t,ha Iletin reduecd to 20 units trvieo daily, 15 gm. of glucose were excreted. Since leaving the hospital on August 11 her carbohydrate tolerance has progressively decreased. In the sixth month of pregnancy, (May 14) she had a positive carbohydrate balanec of 44 gm. 03 30 units of II&in, H-20 strength, available COH 54. On December 7, four and one-half months postpartum, there was a positive balance of 10 gm. on 60 units of H-20 11&n, (the equivalent of 85 units of the H strength), available COH 40. The fetal autopsy revealed certain findings in the pancreas of questionable significance. The gland was normal to gross examination. Microscopic examination showed that the glandular structures mere surrounded by an edematous net-work of eonneetive tissue. Some of the islands of Langerhans were larger than norn1a.l normal. There was ar,d their cells marliedly edematous. Other idands appeared no evidence of the hydropic degeneration typical of pancreatic diabetes.

The local tissue reaction to the lletin injections was more severe in this patient during her pregnancy than it has been postpartum. She objected to increasing the number of injections beyond two a. day, and to increasing the size of the individual dose. Xt is possible that if she had been hospitalized during the last two months of her pregnant term and the larger amount of insulin given which was at times indicated, a living child might have been secured. But it is unlikely that the evident effect of the pregnancy in decreasing the patient’s tolerance would have been modified. The strain imposed by pregnancy upon the general metabolism and upon the endocrine glands is especially severe upon the damaged pancreas of the diabetic, and its effect can apparently not be warded accordingly, the moderately severe or severe diabetic off by insulin. should be prevented from incurring the risk of a pregnancy, and if pregnant an early interruption is indicated. 110 WEST SEVENTY-NINTH STREET.