Nitrogen Metabolism and Insulin
Requirements in Obese Diabetic Adults
on a Protein-Sparing Modified Fast
Bruce R. Bistrian, M.D., Ph.D., George L. Blackburn, M.D., Ph.D.,
Jean-Pierre Flatt, Ph.D., Jack Sizer, M.D.,
Nevin S. Scrimshaw, Ph.D., M.D., and Mindy Sherman, B.A.,
Boston and Cambridge, Massachusetts
SUMMARY
A protein-sparing modified fast (PSMF), which is a total fast
modified by the intake of 1.2-1.4 gm. protein per kilogram ideal
body weight (IBW), fluids ad libitum, and vitamin and mineral
supplementation, allows effective control of carbohydrate
metabolism and hunger. It reduces serum glucose and insulin con-
centrations in obese diabetic patients and increases free fatty acid
and ketone body concentrations; ketonuria appears within 24-72
hours. When this fast was applied to seven obese adult-onset dia-
betics who were receiving 30-100 units of insulin per day, insulin
could be discontinued after 0-19 days (mean, 6.5). In the three
patients who had extensive nitrogen-balance studies, balance could
Adult-onset diabetes mellitus is frequently as-
sociated with obesity,1 and weight loss has repeatedly
been shown to have a beneficial effect on pancreatic
endocrine function.2"5 Unfortunately, the obese sel-
dom achieve weight reduction by diet therapy, irres-
pective of their state of carbohydrate tolerance:6 a
comprehensive review of conventional, outpatient
treatment of obesity revealed that less than 5 per cent
of the patients studied lost 40 pounds or more.7 In the
obese diabetic receiving insulin, the problems of insu-
lin regulation further complicate dieting unless the
dietary composition remains relatively stable from day
to day.
From the Cancer Research Institute, New England Deaconess
Hospital, Harvard Medical School, Boston, Massachusetts 02215,
and the Department of Nutrition and Food Science, Massachusetts
Institute of Technology, Cambridge, Massachusetts 02142.
Address reprint requests to Dr. Bruce R. Bistrian, Cancer Re-
search Institute, 194 Pilgrim Road, Boston, Massachusetts
02215.
Accepted for publication February 18, 1976.
494
be maintained chronically by 1.3 gm. protein per kilogram IBW,
despite the gross caloric inadequacy of the diet. The PSMF was
tolerated well in an outpatient setting after the initial insulin-
withdrawal phase had occurred in the hospital. Significant im-
provements in blood pressure, lipid abnormalities, parameters of
carbohydrate metabolism, and cardiorespiratory symptoms were
associated with weight loss and/or the PSMF. For diabetics with
some endogenous insulin reserve, the PSMF offers significant ad-
vantages for weight reduction, including preservation of lean body
mass (as reflected in nitrogen balance) and withdrawal of exoge-
nous insulin. DIABETES 25:494-504, June, 1976.
Total fasting reduces or eliminates hunger8 and has
proved an effective method of inducing rapid weight
loss.9 Its widespread application has been limited,
however, by significant protein catabolism (represent-
ing nearly half the weight loss in the first month10)
coupled with undesirable physiologic effects (i.e.,
fasting lowers renal creatinine clearance,11 increases
retention of sodium sulfobromophthalein by the
liver,12 raises levels of serum bilirubin,13 reduces con-
centrations of both triiodothyronine14 and urinary
17-hydroxy- and ketosteroids, 15 and induces
neutropenia16).
When a fasting patient is fed small amounts of
protein as egg albumin, 17 casein,18 casein with
glucose,19 or amino acids with glucose,20 the fast-
associated nitrogen loss declines but does not com-
pletely disappear, if one includes estimated integu-
mental nitrogen losses, which was not done in the
previous studies.17"20 When fecal and cutaneous los-
ses are not measured or estimated, the meaning of the
term "positive nitrogen balance" is open to question.
DIABETES, VOL. 2 5 , NO. 6
 BRUCE R. BISTRIAN, M . D . , PH.D., A N D ASSOCIATES

In extensive studies we have found that 1.2-1.4 TABLE l

gm. protein, in the form of egg albumin or lean Anthropometric characteristics of study patients and results of weight
loss in the first year of a protein-sparing modified fast
beef,21 per kilogram ideal body weight (IBW) as an
estimate of lean body mass is sufficient to maintain Initial Ideal
nitrogen balance and to preserve normal liver, endo- Patient Height weight body weight
cm. kg- kg-
crine, and hematopoietic functions in the fasting
1 177.8 152 65.1
obese patient. The beneficial effects of total fasting on 2 152.4 97.5 48.4
carbohydrate metabolism in diabetic patients were 3 160 103 52.6
known even before the discovery of insulin. 22 This 4 180.3 125 71.4
5 165.1 93.6 54.0
modified fast produces a similar fall in serum insulin 6 154.9 71 49.9
and glucose concentrations, a rise in free fatty acids 7 172.7 114 65.8
and ketone bodies, and the appearance of ketonuria. 21 Weight loss on Time required Weight loss at
With the expected fall in insulin requirements com- PSMF (months) 1 year
bined with preservation of nitrogen balance and kg- kg-
1 74 12 74
weight loss, the protein-sparing modified fast (PSMF) 2 28 12 28
was considered appropriate for treatment of obese 3 38 10 28.5
diabetics. Six such patients were monitored carefully 4 19 1 1/2 8.5*
5 18 4 18t
as inpatients while their insulin dosages were gradu- 6 2
9 6t
ally eliminated, and in a seventh, an outpatient, insu- 7 21 4 o§
lin was eliminated. This report details the experience
*20 months.
of these patients and emphasizes the first three cases f 10 months.
because of the extensive nitrogen-balance studies con- |Seven months.
ducted. §17 months.

EXPERIMENTAL PROCEDURE
analyzed for the following: fasting glucose, urea ni-
Subjects and diets. Six diabetics requiring insulin to trogen, electrolytes, magnesium, calcium, phos-
control hyperglycemia, who were obese according to phorus, uric acid, creatinine, total bilirubin, trans-
the Metropolitan Life Insurance Company aminase, alkaline phosphatase, creatine phosphoki-
weight/height/sex standards 23 (table 1), were am- nase, lactic dehydrogenase, total protein, albumin,
bulatory inpatients, and the seventh was an outpa- cholesterol, triglyceride, A.M. and P.M. cortisol, and
tient. Patients 1-3 were maintained on a PSMF sup- total thyroxine. A chest x-ray and electrocardiogram
plying 0.8-1.0 gm. protein (as lean beef or equivalent (ECG) were also obtained.
animal protein) per kilogram body weight (BW) per Body weights, urinary ketones (by Acetest), and
day, at least 1,500 cc. of water, 1 K-Lyte tablet vital signs were recorded daily. Urinalysis, CBC, and
(potassium bicarbonate) (25 mEq.), a One-a-Day Vi- the above assays in blood serum were repeated weekly,
tamin with Iron, two Turns (calcium carbonate, 197 except for the endocrine studies. A substrate
mg. calcium/tablet) twice daily for calcium, and salt profile—which included levels of insulin, free fatty
ad libitum. This regimen supplied 300-750 calories acids, B-hydroxybutyrate, and acetoacetate (measured
per day. Four less severely obese patients (4-7; table 1) by previously described methods 24 )—was obtained
were also fed the PSMF. We had recognized by this approximately every week while the subjects were
time that protein requirements needed to be based on CRC inpatients (table 2). While they were inpatients,
some stable measure of lean body mass, such as IBW, dietary nitrogen content was determined by Kjeldahl
if nitrogen balance was to be achieved; these three analysis of samples of meat bought in large lots in
patients therefore consumed 1.3-1.5 gm. protein per patients 1-3 and estimated from values found in stan-
kilogram IBW, in addition to the vitamin, mineral, dard tables 25 in patients 4-6 during hospitalization
and fluid supplements listed above. and in all patients after discharge. Twenty-four-hour
Protocol. Patients 1-3 were admitted to the Mas- urine samples were collected either continuously or
sachusetts Institute of Technology Clinical Research during three consecutive days each week and were
Center (CRC) and underwent a full clinical evalua- used to measure total nitrogen by the Kjeldahl proce-
tion, including history, physical exam, complete dure. Fecal nitrogen (N) was estimated as the mean
blood count (CBC), and urinalysis. Blood serum was excretion, 0.4 gm. N , calculated from pooled weekly

JUNE, 1976 495

 NITROGEN METABOLISM AND INSULIN REQUIREMENTS

TABLE 2 mained relatively stable, although insulin require-

Free fatty acid and ketone body levels in patients 1-3 ments to regulate hyperglycemia and glycosuria in-
creased to 52 units per day. At the beginning of this
/3-Hydroxy- Aceto- Free fatty study, her weight was 152 kg., height 178 cm.,
butyrate acetate acids blood pressure 225/130 mm. Hg (despite antihyper-
Patient /imoles/ml. ^moles/ml. fiEq./ml.
1: Prefast 0.37 0.01 0.65 tensive medications), glucose 288 mg./lOO ml., tri-
Week 1 0.75 0.00 0.81 glyceride 268 mg./lOO ml., and uric acid 6.9
Week 2 0.93 0.00 1.09 mg./lOO ml. (figures 1-3). The patient began the
Week 3 0.53 0.04 1.45
2: Prefast* — — — PSMF with 150 gm. protein per day; she was given 10
Week 1 0.65 0.14 0.66 units of regular insulin for 3 ~ 4 + glycosuria in urine
Week 2 1.24 0.22 0.70 samples examined four times daily. She received 40
Week 3 1.23 0.31 0.40
3: Prefast 0.11 0.10 0.28 units for each of the first three days, then 20 units for
Week 1 1.45 0.01 1.29 four days, and 10 units for 11 days; after this period
Week 2 2.06 0.00 1.04 (18 days), no further glycosuria appeared and insulin
Week 3 1.34 0.18 0.95
was suspended. As shown in figures 1-3, weight,
*Patient 2 was on a PSMF on admission. blood pressure, glucose, insulin, uric acid, and tri-
glyceride levels fell rapidly with little change in N
collections from eight healthy, obese control patients balance or serum cholesterol. Free fatty acids and
undergoing a similar modified fast. Integumental and ketone bodies increased in serum (table 2), and
miscellaneous N losses were assumed to be 5 mg. N ketonuria appeared within 24 hours. Uric acid levels
per kilogram BW. 2 6 Nitrogen balance was calculated increased initially (an effect related to high serum
according to this formula:
N balance = N intake — (urine N + fecal N 4-
Patient # I 6 I 7 47y.o. $ hgt. 178cm
estimated integumental and miscellaneous N losses)
After discharge, each patient was instructed to fol- _ 300 rr

low the PSMF with weekly or biweekly monitoring

until hospital readmission. During the outpatient vis-
its, weight, urinary ketones, blood pressure, CBC,
and standard serum analysis were measured. The case
reports outline the management of insulin dosage.
Patients 4-6 were admitted to the New England
Deaconess Hospital (NEDH) for a similar full clinical
study, except that substrate profiles and N balance
measurements were not obtained. The outpatient
phase included one or two weekly visits for weight,
urinary ketones, blood pressure, and blood glucose
check, with CBC and standard serum chemistries de-
termined at six-to-eight-week intervals. Patient 7,
with easily regulated diabetes, received the same clin-
ical evaluation but remained an outpatient throughout
the study. In these four patients technics of behavior - 0 -(152)
modification27 and instruction in nutrition education
were included in the treatment program. 3 10
'
^
20 (13 2)
RESULTS
150 I no 1 55 gm Protein/Day
1 i 1
Patient 1 C) 3 16 24 32 40 Days
Patient 1, E.S., was a 47-year-old woman with
morbid obesity beginning in her early teens and a 40ucl eor
family history of diabetes. Diabetes was discovered 11 +I2u h PH

years earlier, when she weighed 141 kg. and required FIG. 1. Acute effects on selected biochemical and clinical
25-30 units of insulin per day. Her condition re- parameters of a PSMF in patient 1 .

496 DIABETES, VOL. 25, NO. 6

 BRUCE R. BISTRIAN, M.D., PH.D., AND ASSOCIATES

# Patient #1617 47y.o. J hgt. 178 cm

Patient 1617 47y.o. J hgt. 178 cm
250 r
- 14
o
Q
•»«.
12
3 10

E
o
O 270-^
o
- 230

S£ 190
o

' 150
A

230

190
-6*
gm Protein/Doy
I 1 '50 \ 0
[ I
2 12 Months
£.2 t
40u Clear
. f no +I2UNPH

i FIG..3. Long-term effects on selected biochemical and clinical

70
I
16 24 32
Days
40u clear
•H2u NPH
FIG. 2. Acute effects on selected biochemical and clinical
parameters of a PSMF in patient 1.
ketones 28 ), but fell after three months despite
maintenance of the PSMF. N balance remained posi-
tive when the patient consumed 150, 110, and 100
gm. protein per day but became negative when con-
sumption fell to 55 gm. protein per day.
Patient 1 was discharged after 10 weeks on 1 gm.
protein per kilogram BW (550-600 calories) and was
followed at 1- or 2-week intervals as an outpatient.
One month later, she was readmitted for two months
to confirm N balance, which remained positive on
parameters of a PSMF in patient 1. Mean N balance ±
standard deviation for three-to-seven-day periods.
40
100 gm. protein per day (figure 3). A 30-day admis-
sion after one year revealed a mildly negative N bal-
ance at intakes of 45 and 60 gm. protein per day
(figure 3). She lost 74 kg. during the 12 months the
PSMF was consumed. Serum glucose, monitored
weekly for the first six months and biweekly thereaf-
ter, remained less than 100 mg./lOO ml. despite only
fair adherence to the diet as an outpatient.
Patient 2
Patient 2, F.F., was a 63-year-old woman with
obesity beginning after age 20; her diabetes, initially
requiring oral agents, had been diagnosed six years
earlier. After four years of clinical diabetes, she suf-
fered a cardiac arrest during an acute asthmatic attack,
after which 50 units of NPH insulin daily was pre-
scribed. On admission to the NEDH because of poor
diabetic control, the dosage had increased to 100 units
of NPH insulin per day. Additional medical problems

JUNE, 1976 497

 NITROGEN METABOLISM AND INSULIN REQUIREMENTS
included atherosclerotic heart disease, chronic ob-
structive pulmonary disease, and asthma. Clinical
data on admission were weight 97.5 kg.; height
152.5 cm.; blood pressure 140/80 mm. Hg; wheezes
in both lung fields; and serum glucose 173 mg./lOO
ml. A chest x-ray showed left ventricular predomi-
nance, ECG revealed possible pulmonary disease, and
pulmonary function tests substantiated the presence of
severe, obstructive pulmonary disease. The PSMF
began at 75 gm. of protein per day, with 25 gm. of
carbohydrate added for five days; NPH insulin was
initially reduced to 60 units in the morning and 20
units in the evening, further lowered to 40 units after
nine days, and discontinued after 10 more days, with
serum glucose remaining in the high 100s range. The
patient lost 7 kg. in the three weeks of admission. Her
serum glucose following discharge was noted to be in
poor control because of deviations from the PSMF, so
she was started on oral agents (tolbutamide, 1.5 gm.
daily).
Three months later she was readmitted to the CRC
weighing 83 kg., with serum glucose 207 mg./lOO
ml., uric acid 9.6 mg./lOO ml., cholesterol 293
mg./lOO ml., and triglyceride 174 mg./lOO ml. The
PSMF continued at 65 gm. protein per day; serum
glucose dropped to 140 mg./lOO ml. after 24 hours
and to below 120 mg./lOO ml. three days later (figure
4). Tolbutamide was discontinued after six days, and
subsequent glucose levels were 80-110 mg./lOO ml.
during this 18-day admission. N balance was positive
on the eighth day (figure 4). As expected, glucose,
insulin, and triglyceride levels fell while uric acid
concentrations increased (uric acid subsequently re-
turned to normal levels despite continued fasting (fig-
ure 5)). Ketonemia (table 2) and ketonuria were per-
sistent. Although patient 2 did not adhere closely to
the diet in the next four months as an outpatient, her
serum glucose remained less than 140 mg./lOO ml.
off medications. On readmission at one year, her
weight was 68 kg., serum glucose 109 mg./lOO ml.,
uric acid 6.2 mg./lOO ml., cholesterol 222 mg./lOO
ml., and triglyceride 134 mg./lOO ml. N balance was
minimally negative on 55 gm. protein per day (0.8
gm. per kilogram BW). Glucose levels were subse-
quently less than 100 mg./lOO ml. during the two-
week admission (figure 5).
Patient 3
Patient 3, S.G., was a 60-year-old woman, obese
since age 26, with a family history of diabetes. Dia-
betes had been diagnosed 11 years earlier and was
managed initially with oral agents, then with insulin
after she had suffered a transient, drug-induced renal
498
Patient # I 8 6 5 63y.o. J hgt. 152.5cm
200
140 -
s8'6o
100 ~
o E
120
60
?i
80 o
20
+2
0
-4
(83)
0
2
4
6 (78)
65gm Protein/Oay
16
Days
1.5 gm Orinose
FIG. 4. Acute and long-term effects on selected biochemical and
clinical parameters of a PSMF in patient 2. Mean N bal-
ance ± standard deviation for three-to-seven-day periods.
failure. The patient received 55 units of lente insulin
daily to regulate hyperglycemia, although diabetic
control was poor even with this dose. Additional med-
ical problems were atherosclerotic heart disease (with a
previous myocardial infarction and angina), a previous
parathyroidectomy accompanied by mild hypothy-
roidism, gout, and a previous hysterectomy and
cholecystectomy. Her clinical data upon admission to
the Thorndike Clinical Research Center (Boston City
Hospital) were weight 103 kg.; height 160 cm.;
blood pressure 120/60 mm. Hg; mild peripheral
neuropathy; serum glucose 178 mg./lOO ml.; tri-
glyceride 363 mg./lOO ml.; uric acid 4.7 mg./lOO
ml.; and thyroxine 4.5 /ug/100 ml.; ECG normal. A
chest x-ray revealed left ventricular hypertrophy. She
initially began the PSMF at 99 gm. protein per day (1
gm. per kilogram BW), and insulin was discontinued
simultaneously. Blood sugar levels remained normal
and N balance was positive (figure 6). She was dis-
charged after 15 days (and a 3-kg. weight loss) and
then readmitted to the CRC for brief periods every
month to monitor the PSMF and to confirm N balance
(figure 7), which remained essentially positive at 80,
75, 70, and 65 gm. of protein. Glucose, insulin, and
DIABETES, VOL. 25, NO. 6
 BRUCE R. BISTRIAN, M.D., PH.D., AND ASSOCIATES

Patient #1865 63y.o. J hgt. 152.5cm Patient # I676 6Oy.o. J hgt. 160 cm
200 7 0
~
= 140 E

T= 160 : 8 120 50 S
in O
S o lioo
- "X 120 30 J
80 o
~~ — — — <
80 10

10
*8
5 p"> 8
\
0
6
(99)
300 ~ o
•Si . 6 o

2^^200 -1 2
(96)
•
I
o
I
* 4 9 9 gm Protein /Day
I I
• o
100 |0 2 10 12
55u Lente Days
Or-. 04) +3
FIG. 6. Acute and long-term effects on selected biochemical and
^ 10 K clinical parameters of a PSMF in patient 3.
to
to
o -J-V-f
years earlier, required insulin for control of carbohy-
^ 20
drate metabolism; by the time of admission, dosage
30 had been increased to 30 units of lente insulin per day.
75 55 gm Protein/Doy
I 1 1 I
Other significant medical problems included severe
0 5 6 10 12 14 Months hypertension, atherosclerotic heart disease, and gout.
On admission his weight was 125 kg., height 182
FIG. 5. Acute and long-term effects on selected biochemical and cm., blood pressure 190/120 mm. Hg (with an-
clinical parameters of a PSMF in patient 2. Mean N bal-
ance ± standard deviation for three-to-seven-day periods. tihypertensive medications), serum glucose 138
mg./lOO ml., and uric acid 6.9 mg./lOO ml. (with
triglyceride dropped (figures 6 and 7); free fatty acids allopurinol). His chest x-ray revealed an enlarged
and serum and urinary ketones increased (table 2); and heart, and ECG showed nonspecific changes. The
uric acid initially rose, then later fell (figure 7)—all PSMF began at 125 gm. protein per day; insulin was
routine effects of the PSMF.21 Angina had prevented initially reduced to 20 units per day and was discon-
earlier replacement of thyroid hormone. When tinued after four days, with glucose levels remaining
thyroxine fell further, to 4.0 /xg/100 ml. with the in the 90-110 mg./lOO ml. range. Ketonuria ap-
development of hypercholesterolemia, hyporeflexia, peared on the third day. The patient lost 10 kg. in 18
and low ECG voltage, the patient was started on a days of hospitalization, and his blood pressure fell to
slow administration of triiodothyronine, which re- the 150/90-100 range on reduced dosage of antihyper-
duced the cholesterol levels. Weight loss in the 10 tensive medication.
months of a PSMF, six weeks as an inpatient, was 38 Discharged, the patient could not follow the diet
kg., with 28.5 kg. weight loss at one year. and regained 5 kg. in 10 weeks. He developed
polyuria and polydipsia and returned to 20 units of
Patient 4 lente insulin a day. In order to ease the operative
Patient 4, C.V., was a 55-year-old man with a closure of his colostomy, he was readmitted for weight
family history of diabetes in whom mild obesity had reduction; clinical data were weight 121 kg.; blood
developed two years earlier, after a colostomy for an pressure 130/94 mm. Hg (on medication); urine glu-
abdominal gunshot wound. Diabetes, diagnosed five cose, trace; serum glucose 172 mg./lOO ml.; and uric
JUNE, 1976 499
 NITROGEN METABOLISM AND INSULIN REQUIREMENTS
Patient * 1676 6Oy.o. £ hgt. 160 cm
1991901 80 75 70 | gm Protein/Doy
0 10 Months
t I Cytomel or I
1 Synthroid |
55u Lente
FIG. 7. Acute and long-term effects on selected biochemical and
clinical parameters of a PSMF in patient 3.
Mean N balance ± standard deviation for three-to-
seven-day periods.
acid 8.2 mg./lOO ml. The PSMF began at 120 mg.
protein per day, and insulin was simultaneously dis-
continued. In six weeks he lost 19 kg.; his blood
pressure returned to normal without medication, and
glucose levels remained normal. Two weeks before
discharge, a reanastomosis of his transverse colon was
performed. The patient was again unable to follow a
PSMF as an outpatient and returned to the use of
insulin. Weight 20 months after his first admission
was \\6Vi kg., SVi kg. less than initially.
Patient 5
Patient 5, E.C., was a 49-year-old woman, obese
500
since childhood, with diabetes of eight years' duration
initially controlled with chlorpropamide and phen-
formin; daily insulin (10 units regular, 90 units NPH
to control hyperglycemia) was required after five
years. Additional medical problems included
peripheral neuropathy, hypertension, and type IV
hyperlipidemia. Clinical data on admission indicated
weight 93.6 kg., height 165.1 cm., blood pressure
160/100 (on medication); serum glucose 222 mg./lOO
ml.; triglyceride 366 mg./lOO ml. She began on 73
gm. protein per day; 16, 10, 8 units NPH insulin
each day were the single weaning doses with blood
glucose in good control. Antihypertensive medica-
tions were discontinued with blood pressure returning
to normal. Discharged after one week, she lost 18 kg.
in four months, with excellent diabetic control off
insulin and maintenance of this weight in the follow-
ing six months.
Patient 6
Patient 6, B.S., was a 60-year-old woman, obese
since childhood, with diabetes of 23 years' duration.
Initially controlled with tolbutamide, daily insulin
(25-30 units of NPH, to regulate carbohydrate
metabolism) was required after 10 years. Additional
medical problems included a previous cholecystec-
tomy with postoperative hepatitis, hypertension, and
a previous myocardial infarction. Clinical data on ad-
mission indicated weight 71 kg.; height 155 cm.;
blood pressure 120/80 mm. Hg (on medication);
bilateral retinal microaneurysms with exudates in the
left eye; diminished pedal pulses; serum glucose 216
mg./lOO ml.; cholesterol 327 mg./lOO ml.; and tri-
glyceride 210 mg./lOO ml. ECG revealed left axis
deviation as well as an old anterior and inferior
myocardial infarction. She began on 80 gm. protein
per day; insulin and the antihypertensivje_drug were
discontinued. After 24 hours, the fasting serum glu-
cose dropped to the 130-to-140-mg./l00 ml. range
and blood pressure was normal. Discharged after five
days, she had lost 9 kg. in the two-month period and
showed both persistent ketonuria and good diabetic
control without insulin, with weight regain of only
0.3 kg. in the following five months.
Patient 7
Patient 7, M.R., a 52-year-old man, had a family
history of diabetes and had been obese since early
adulthood. Diabetes had been present for 15 years and
required a daily dose of 40 units lente and 20 units
regular insulin to regulate carbohydrate metabolism.
Additional medical problems included a history of
multiple renal stones and recurrent pyelonephritis,
DIABETES, VOL. 2 5 , NO. 6
 BRUCE R. BISTRIAN, M.D., PH.D., AND ASSOCIATES
chronic obstructive pulmonary disease, carcinoma of
the prostate, and hernias at the site of a cholecystec-
tomy scar and in the umbilical area. Clinical data were
weight 114 kg.; height 172.7 cm.; blood pressure
150/90 mm. Hg; exudate in the right eye; serum
glucose 272 mg./lOO ml. An ECG revealed an old
inferior myocardial infarction. The PSMF began with
110 gm. protein per day on an outpatient basis be-
cause he refused hospital admission. Insulin dosage
was initially reduced by 50 per cent and was discon-
tinued after one week. In four months he lost 21 kg.
and had a blood pressure of 135/80 mm. Hg. His
adherence to the fast was poor, however, with exces-
sive protein intake; his serum glucose levels were in
the high, 100-200 mg./lOO ml. range. For this
reason, a daily dose of 10-15 units N P H was begun
after two months and produced better diabetic con-
trol. Weight 13 months later was identical to initial
weight.
SUMMARY
The patients tolerated the dietary protocol well.
Some patients developed mild symptoms of postural
hypotension, which generally subsided after seven to
10 days. Both systolic and diastolic blood pressures
fell significantly, often to normal levels, in the hyper-
tensive patients. Pulse rates and temperatures tended
to be low. Bowel movements were less frequent, and
patients occasionally required mild cathartics. No
problems were encountered aside from occasional in-
tercurrent, upper respiratory illness. The three CRC
patients with significant cardiorespiratory symptoms
all improved dramatically.
No significant biochemical changes were noted in
the standard clinical chemistries except for a transient
lowering of blood urea nitrogen; serum glucose and
insulin, levels dropped, while free fatty acid and
ketone body levels rose. Similar effects of a PSMF on
normal o b e s e 2 0 2 4 2 9 3 0 and on nonobese patients 3 1 3 2
have been reported. (It should be noted, however, that
well-recognized difficulties are associated with the
measurement of insulin by radioimmunoassay after
exogenous insulin has been administered. 3 3 )
Ketonuria occurred within 24 to 72 hours and per-
sisted when patients adhered to the regimen. Serum
uric acid levels initially rose beyond the normal range,
then fell to baseline and below after several months,
despite continuation of the modified fast. Serum
cholesterol and triglyceride dropped in all patients,
with a late transient increase of cholesterol in patients
1 and 3 (in the latter case being associated with
hypothyroidism). Nitrogen balance was obtained
JUNE, 1976
when at least 1.3 gm. protein per kilogram IBW were
provided daily to the CRC patients.
With these seven patients who needed insulin to
manage hyperglycemia and glycosuria when consum-
ing an ad libitum diet, insulin could be discontinued
after 0-19 days (mean, 6.5 days) of modified fasting
and could have been withdrawn sooner in patients 1
and 2. Urinary glycosuria in conjunction with serum
glucose level was used to determine the insulin re-
quirements during the weaning phase, following
which excellent control of carbohydrate metabolism
was achieved.
Five of the seven patients sustained weight loss of
greater than 40 pounds primarily as outpatients. At
intervals of 12 months (three patients) and 5-10
months (three patients) substantial weight loss was
maintained. One patient (no. 4), unsuccessful at out-
patient reduction, still maintained some weight loss
at 20 months (table 1).
DISCUSSION
Hyperinsulinism occurs in both the normal and
diabetic obese, and weight loss improves carbohydrate
tolerance and lowers serum insulin in both. 3 5 3 6 Ele-
vated insulin levels are clearly related to increased
body fat,4 in part because certain tissues, such as
muscle 37 ; 38 and adipose, 39 are resistant to insulin and
in part because the fat mass is increased.40 Because it
is the rate of free fatty acid release per unit mass of
adipose tissue, rather than the production of free fatty
acid, which is regulated by the concentration of insu-
lin, the severalfold increase in adipose tissue mass as-
sociated with obesity results in elevation of free fatty
acid levels40 and hyperinsulinism. 41 If one considers
the metabolic situation in the obese, without having
to make allowance for an impaired responsiveness of
adipose tissue to insulin, an interesting consequence
becomes apparent. Hyperinsulinism will tend to es-
tablish a positive caloric balance in adipose tissue, as
insulin favors glucose and triglyceride uptake while
inhibiting mobilization of fat. The realization that
hyperinsulinism converts adipose tissue into a sink for
calories implies that weight-reducing regimens should
be specifically designed to allow a sharp fall in insulin
levels. Thus, avoidance of carbohydrates assumes spe-
cial importance in the fasting obese, given this trend
toward hyperinsulinism not only because of the
anabolic effect of insulin on adipose tissue but because
hyperinsulinemia per se produces insulin resistance.42
It is important to recognize that the starvation
ketosis that results on fasting or modified fasting does
 NITROGEN METABOLISM AND INSULIN REQUIREMENTS
not induce the pathologic ketoacidosis that occurs in
decompensated diabetes. Ketogenesis is regulated by
insulin, 43 and any marked increase in ketone body
levels stimulates insulin secretion to establish a feed-
back control. 44 These increased ketone body levels are
important since evidence has been presented that they
are largely responsible for the protein-sparing
effect. 2 4 3 0 4 5 Not only do ketones substitute for
protein-derived glucose as a fuel for brain, but their
infusion during fasting further conserves nitrogen. 45
This latter ketone-induced protein conservation may
be due to reduced branched-chain amino acid (bcaa)
oxidation in muscle, 45 since physiologic increases in
betahydroxybutyrate inhibit bcaa oxidation in
muscle, 46 which in turn decreases the rate of over-all
catabolism in muscle. 47 Ketone body levels are higher
in total 48 than in a PSMF, 24 but without protein
intake, maximal nitrogen sparing is —5 gm. nitrogen
per day late in total starvation. With modified fasting
the presence of exogenous protein allows better net
protein sparing than fasting, with further improve-
ment as ketone bodies increase in the serum. Technics
to modify a fast that incorporate car-
bohydrate 1920 and thus cause insulin secretion can be
expected to lower ketone body levels and net protein
sparing at any given total calorie intake in the semi-
starvation range.
A comparison of the efficacy of protein sparing in
semistarvation by total fasting, the PSMF, or mod-
ified fasting is possible. Total fasting, when fully ef-
fective after about three weeks, results in a daily loss
of approximately 0.3 kg. fat and 0.1 kg. lean tissue
(in men) or 0.2 kg. fat and 0.1 kg. lean tissue (in
women). 49 This net protein catabolism of acaloric
dieting—with adverse impact on organ function
—makes it undesirable for outpatient use. Total
weight loss in the PSMF is 0.3 and 0.2 kg. of fat per
day for males and females, respectively,21 with nitro-
gen balance maintained, suggesting nearly complete
preservation of lean body mass. In fasting (or modified
fasting), estimates of changes in lean body mass using
nitrogen balance are more accurate than the usually
more exact technics of serial measurements of whole-
body potassium, body density determined by under-
water weighing, and total body water due to the water
and potassium shifts seen with starvation. 50
During modified fasting on fixed and smaller
amounts of protein than the PSMF with* 1 9 2 0 or
•These two dietary regimens 19 ' 20 are considered to be modified
fasting despite the presence of glucose, because significant ketosis
develops.
without 1 7 1 8 some added carbohydrate, considerable
endogenous protein is lost. The failure to consider
nitrogen losses from stool and integument by three of
these authors 18 " 20 leads to overestimation of nitrogen
balance by approximately 1 gm. nitrogen/day, equi-
valent to 6.25 gm. of protein and 25 gm. of lean
tissue. 49 While such losses are not important over the
short term, a significant impact on body cell mass and
function occurs when such a regimen is pursued for
many months. Further, in three of these reports 1719 - 20
balance studies were conducted immediately follow-
ing a total fast, which will transiently improve effi-
ciency of dietary protein utilization. 51 As another
source of nitrogen imbalance in some subjects, it is
unlikely that one fixed level of protein intake can be
equally effective in sparing body protein in semistar-
vation without regard to the primary variables for
amount of lean body mass—sex, age, height, and
weight. The combined effect of the true nitrogen loss
may present as hair loss19 or, more importantly, as
anemia or leuko- or neutropenia. 20 Whatever the
amount of lean body mass lost, the ultimate return to
a maintenance diet containing recommended daily al-
lowances for protein and calories must result in some
weight regain due to the restoration of this deficit of
lean tissue.
An alternative means of weight reduction,
balanced-deficit dieting, has been generally unsuccess-
ful for significant weight loss in nondiabetics7 and
diabetics, 6 who may also have the added difficulty of
regulating insulin doses as the caloric content of the
diet varies. Furthermore, significant protein
catabolism also occurs during substantial weight loss
on hypocaloric, mixed diets. 52 " 54
Thus, for diabetics with some endogenous insulin,
the PSMF not only leads to best maintenance of lean
body mass during weight reduction but, like the other
fasting technics, also allows the early withdrawal of
exogenous insulin. No controlled study was con-
ducted to determine the optimal schedule for discon-
tinuing insulin, but one to two weeks at half the
customary dose would appear sufficient. Insulin can
be withdrawn immediately from obese patients requir-
ing only 30-35 units a day. Until further experience is
gained, this regimen should be initiated only with
inpatients.
ACKNOWLEDGMENTS
The authors acknowledge the help of the nursing,
dietary, and laboratory staff at the Massachusetts In-
stitute of Technology Clinical Research Center. This
DIABETES, VOL. 2 5 , NO. 6
 BRUCE R. BISTRIAN, M.D., PH.D., AND ASSOCIATES
study was supported in part by MIT Training Grant
TO1-AM-05371 (Dr. Bistrian), by MIT Clinical Re-
search Center MO1-RR 00088 (Dr. Bistrian) and
Grant GM 8981-2 from the Cancer Research Insti-
tute, New England Deaconess Hospital (Dr. Black-
burn and Dr. Bistrian).
REFERENCES
'Joslin, E.P., Dublin, L.I., and Marks, H.H.: Studies in dia-
betes mellitus. IV. Etiology. Am. J. Med. Sci. 292:9-23, 1936.
2
Karam, J.H., Grodsky, G.M., and Forsham, P.H.: Excessive
insulin response to glucose in obese subjects as measured by im-
munochemical assay. Diabetes 22:197-204, 1963.
3
Kalkhoff, R., and Ferrou, C : Metabolic differences between
obese overweight and muscular overweight men. N. Engl. J.
Med. 284:1236-39, 1971.
4 25
E1-Khodary, A.Z., Ball, M.F., Oweiss, I.M., and Canary,
J.J.: Insulin secretion and body composition in obesity.
Metabolism 22:641-55, 1972.
5
Olefsky, J.M., Reaven, G.M., and Farquhar, J.W.: Effects of
weight reduction on obesity. Studies of lipid and carbohydrate
metabolism in normal and hyperlipoproteinemic subjects. J. Clin.
Invest. .53:64-76, 1974.
6
West, K.M.: Diet therapy of diabetes: An analysis of failure.
Ann. Internal Med. 79:425-34, 1974.
7
Stunkard, A., and McLaren-Hume, M.: Results of treatment
for obesity: Review of literature and report of series. Arch. Intern.
Med. 203:79-85, 1959.
8
Bloom, W.L.: Fasting as an introduction to the treatment of
obesity. Metabolism 8:214-20, 1959.
9 30
Thompson, T.J., Runcie, J., and Miller, V.: Treatment of
obesity by total fasting for up to 249 days. Lancet 2:992-96,
1966.
10
Ball, M.F., Canary, J.J., and Kyle, L.H.: Comparative ef-
fects of caloric restriction and total starvation on body composition
and obesity. Ann. Intern. Med. 67:60-67, 1967.
n 32
Edgren, B., and Wester, P.O.: Impairment of glomerular
filtration in fasting for obesity. Acta Med. Scand. 290:389-93,
1971.
12
Rozental, P., Biava, C , Spencer, H., and Zimmerman,
H.J.: Liver morphology and function tests in obesity and during
total starvation. Am. J. Digest. Dis. 22:198-208, 1968.
13
Barrett, P.V.: Hyperbilirubinemia of fasting. J. A. M. A.
227:1349-53, 1971.
14
Portnay, G.I., O'Brian, J.T., Bush, J., Vagenakis, A.G.,
Azizi, F., Arky, R.A., Ingbar, S.H., and Braverman, L.E.: The
effect of starvation on the concentration and binding of thyroxine
and triiodothyronine in serum and on the response to TRH. J.
Clin. Endocrinol. Metab. 39:191-94, 1974.
15
Schachner, S.H., Wieland, R.G., Maynard, D.E., Kruger,
F.A., and Hamwi, G.J.: Alterations in adrenocortical function in
fasting obese adults. Metabolism 24:1051-58, 1968.
16
Drenick, E.J., and Alvarez, L.C.: Neutropenia in prolonged
fasting. Am. J. Clin. Nutr. 24:859-63, 1971.
17
Bollinger, R.E., Lukert, B.P., Brown, R.W., Guevara, L.,
and Steinberg, R.: Metabolic balance of obese subjects during
fasting. Arch. Intern. Med. 2 28:3-8, 1966.
18
Apfelbaum M., Boudon, P., Lacatis, D., and Nillus, P.:
Effets metaboliques de la diete protidique chez 41 sujets obeses.
Presse Med. 78:1917-20, 1970.
JUNE, 1976
19
Genuth, S.M., Castro, J.H., and Vertes, V.: Weight reduc-
tion in obesity by outpatient semistarvation. J. A. M. A.
230:987-91, 1974.
20
Baird, I.M., Parsons, R.L., and Howard, A.N.: Clinical and
metabolic studies of chemically defined diets in the management
of obesity. Metabolism 23:645-57, 1974.
"Blackburn, G.L., Bistrian, B.R., Flatt, J.P., and Sizer, J.:
Role of a protein sparing modified fast in a comprehensive weight
reduction program. In Recent Advances in Obesity Research: I,
Howard, Alan, Ed. London, Newman Publishing Ltd., 1975.
"Allen, F.M., Stillman, E., and Fitz, R.: Total dietary regula-
tion in the treatment of diabetes. Monograph 11. New York,
Rockefeller Institute of Medical Research, 1919, p. 90-97.
2
'Society of Actuaries: Build and blood pressure study, Vols. I
and II. Chicago, Society of Actuaries, 1959.
"Blackburn, G.L., Flatt, J.P., Clowes, G.H.A., O'Donnell,
T.F., and Hensle, T.E.: Protein-sparing therapy during periods of
starvation with sepsis or trauma. Ann. Surg. 277:588-94, 1973.
Nutritive Value of Foods. Home and Garden Bulletin No.
72. Washington, U.S. Government Printing Office, 1970.
26
Joint FAO/WHO Expert Committee: Energy and Protein
Requirements. WHO Tech. Rep. Series No. 522. Geneva, World
Health Organization, 1973.
"Stuart, R.B., and Davis, B.: Slim Chance in a Fat World,
Behavioral Control of Obesity. Champaign, 111., Research Press,
1972.
28
Lecocq, F.R., and McPhaul, J.J.: The effects of starvation,
high fat diets and ketone infusions on uric acid balance.
Metabolism 24:186-97, 1965.
29
Weisswange, A., Cuendet, G.S., and Marliss, E.B.:
Metabolic response to a "revolutionary" diet (abstr.). Clin. Res.
22:482A, 1974.
Marliss, E.B., and Nakhooda, A.F.: Ketosis and protein-
sparing in man (abstr.). Clin. Res. 22:751A, 1974.
31
Freeman, J.E., Stegink, L.D., Fry, L.K., Sherman, B.M.,
and Denbesten, L.: Evaluation of amino acid infusions as protein-
sparing agents in normal adult subjects. Am. J. Clin. Nutr.
28:477-81, 1975.
Hoover, H.C., Grant, J.P., Gorschboth, C , and Ketcham,
A.S.: Nitrogen-sparing intravenous fluids in postoperative pa-
tients. N. Engl. J. Med. 293:172-75, 1975.
33
Isles, T.E., and Farquhar, J.W.: The effect of endogenous
antibody on insulin assay in the newborn infants of diabetic
mothers. Pediatr. Res. 2:110-15, 1967.
34
Genuth, S.M.: Plasma insulin and glucose profiles in normal,
obese and diabetic persons. Ann. Intern. Med. 79:812-22, 1973-
35
Yalow, R.S., Glick, S.M., Roth, J., and Berson, S.A.:
Plasma insulin and growth hormone levels in obesity and diabetes.
Ann. N.Y. Acad. Sci. 232:357-73, 1965.
36
Karam, J.H., Grodsky, G.M., Parlatos, F.C., and Forsham,
P.H.: Critical factors in excessive insulin response to glucose.
Lancet 2:286-89, 1965.
37
Rabinowitz, D., and Zierler, K.L.: Forearm metabolism in
obesity and its response to intra-arterial insulin. Characterization
of insulin resistance and evidence for adaptive hyperinsulinism. J.
Clin. Invest. 42:2173-81, 1962.
38
Butterfield, W.J.H., Hanley, T., and Whichelow, M.J.:
Peripheral metabolism of glucose and free fatty acids during oral
glucose tolerance tests. Metabolism 24:851-66, 1965.
39
Salans, L.B., Knittle, J.L., andHirsh, J.: The role of adipose
cell size and adipose tissue sensitivity in the carbohydrate intoler-
ance of human obesity. J. Clin. Invest. 47:153-65, 1968.
 NITROGEN METABOUSM AND INSUUN REQUIREMENTS
40
Flatt, J.P.: Role of the increased adipose tissue mass in the
apparent insulin insensitivity of obesity. Am. J. Clin. Nutr.
2.5:1189-92, 1972.
41
Randle, P.J., Garland, P.B., Hales, C.N., and Newsholme,
E.A.: The glucose fatty-acid cycle: Its role in insulin sensitivity
and the metabolic disturbances of diabetes mellitus. Lancet
2:785-89, 1963.
42
Gavin, J.R., Roth, J., Neville, D.M., DeMeyts, P., and
Buell, D.N.: Insulin-dependent regulation of insulin receptor
concentrations: A direct demonstration in cell culture. Proc. Natl.
Acad. Sci. U.S.A. 72:84-88, 1974.
43
Wieland, O.: Ketogenesis and its regulation. Adv. Metab.
Disord. 3:1-47, 1968.
44
Madison, L., Mebane, D., Unger, R.H., and Lochner, A.:
The hypoglycemic action of ketones. II. Evidence for stimulatory
feedback of ketones in the pancreatic beta cells. J. Clin. Invest.
43:408-15, 1964.
45
Sherwin, R.S., Hendler, R.G., and Felig, P.: Effect of
ketone infusions on amino acid and nitrogen metabolism in man.
J. Clin. Invest. .5.5:1382-90, 1975.
46
Buse, M.G., Biggers, J.F., Friderici, K.H., and Buse, J.F.:
Oxidation of branched chain amino acids by isolated hearts and
diaphragm of the rat. The effect of fatty acids, glucose, and pyru-
vate respiration. J. Biol. Chem. 247:8085-96, 1972.
47
Fulks, R.M., Li, J.B., and Goldberg, A.L.: Effects of insu-
lin, glucose, and amino acids on protein turnover in rat dia-
phragm. J. Biol. Chem. 250:290-98, 1975.
48
Owen, O.E., Felig, P., Morgan, A.P., Wahren, J., and
Cahill, G.F., Jr.: Liverand kidney metabolism during prolonged
starvation. J. Clin. Invest. 48:574-83, 1969.
49
Runcie, J., and Hilditch, T.E.: Energy provision, tissue
utilization, and weight loss in prolonged starvation. Br. Med. J.
2:352-56, 1974.
50
Drenick, E.J.: Exogenous thyroid hormones to accelerate
weight loss. Obesity Bariatric Med. 4:244, 1975.
"Bistrian, B.R., Blackburn, G.L., Flatt, J.P., Winterer, J.,
Cochran, D., and Young, V.: Protein requirements for net
anabolism with a hypocaloric diet. Clin. Res. (Abstr.) 23:315A,
1975.
52
Quinn, M., Kleeman, C.R., Bass, D.E., and Henschel, A.:
Nitrogen, water and electrolyte metabolism in protein and
protein-free low calorie diets in man. II Adequate water intake.
Metabolism 3:68-77, 1953.
53
Keyes, A., Brozek, J., Henschel, A., Michelson, O., and
Taylor, H.: The Biology of Human Starvation, Vols. I and II.
Minneapolis, University of Minnesota Press, 1950.
54
Drenick, E.J.: Weight reduction with low calorie diets; Prac-
tical management. J. A. M. A. 202:118-20, 1967.
DIABETES, VOL. 2 5 , NO. 6