Influence of Various Sugars, Chromium and Other
Trace Metals on Serum Cholesterol
and Glucose of Rats '
HENRY
A. SCHROEDER,2
Department
Hampshire
of Physiology,
and Brattleboro
MARIAN
Dartmouth
Memorial
MITCHENER
Medical
Hospital,
AND ALEXIS
P. NASON
School, Hanover,
New
Brattleboro,
Vermont
The cholesterol-lowering effect of chro
mium and of brown sugar fed to rats on
a torula yeast—sucrose—larddiet has been
reported (1). Rats given refined white
sugar had higher serum levels of choles
terol than did those given added chromium
(III) by 11 months of age, and higher
levels than rats fed dark brown sugar by
5 and 11 months of age. Others fed raw
sugar up to 4 months of age had somewhat
depressed serum cholesterol levels. The
observations on this effect of chromium
have been confirmed in hypercholesteremic
rats of another strain fed cholesterol (2);
chromium suppressed elevated serum lev
els significantly. Fasting serum glucose was
also depressed by brown sugar (1).
We have extended these observations to
22 to 23 months of age, have examined the
J. NUTRITION, 101: 247-258.
effects of chromium, nickel, molybdenum
and cadmium on fasting serum cholesterol
levels of rats on a starch diet, and investi
gated glycolytic effects of these trace ele
ments. Although a part of the data in this
study has been reported (1), they are in
cluded for purposes of ascertaining changes
with age.
METHODS
Two diets were used in this study. A
diet of 30% torula yeast,3 50% sugar, 15%
Received for publication May 22, 1970.
1 Supported by Public Health Service Research Grant
no. HE 05076 from the National Heart Institute; CIBA
Pharmaceutical
Products Inc.; Cooper Laboratories,
Inc.; and General Foods Corporation.
2 Present address: 9 Belmont Avenue, Brattleboro,
Vermont 05301.
3 General Biochemicals, Inc., Chagrin Falls, Ohio.
247
Downloaded from jn.nutrition.org by on November 25, 2008
ABSTRACT
Because the feeding of brown sugar or chromium(III)
to rats
had been found to produce lowered serum cholesterol levels for 11 months,
studies were continued to 23 months of age to ascertain long-term effects. In
addition an attempt was made to discover other trace factors influencing choles
terol and glucose metabolism. Groups of rats were fed a torula yeast, sucrose and
lard diet, with refined white sugar, dark brown sugar or raw sugar. The diet con
tained 0.08, 0.16 and 0.14 Mg/g chromium, respectively, dry weight. One group
fed white sugar was given 5 ppm chromium in drinking water, and all were given
the other essential trace metals. A standard diet of rye, milk and corn oil was fed
to other groups of rats, with or without chromium in water, and with cadmium,
nickel and molybdenum. The feeding of white sugar plus chromium, brown sugar
or raw sugar retarded the rise of serum cholesterol with age characteristic of
chromium deficiency. The feeding of brown sugar lowered fasting serum glucose
levels up to 11 months of age but not thereafter in females; in males the level
rose after 11 months but did not rise to the levels of the white sugar group. Chro
mium was hyperglycémiewith age, whereas the feeding of brown sugar to male
rats was accompanied by lesser elevations at older ages, and raw sugar had mini
mal effects in both sexes. In rats fed the starch diet, the addition of 5 ppm chro
mium usually resulted in depression of serum cholesterol and glucose levels, the
addition of 50 ppm cadmium had no consistent effects, and the feeding of 50 ppm
molybdenum alone without other trace elements resulted in effects on cholesterol
similar to those of chromium. The torula yeast diet contained little vitamin E,
and neurological symptoms of vitamin E deficiency occurred in all animals fed
this diet for 18 to 24 months.
J. Nutr. 101: 247-258, 1971.
248
H. A. SCHROEDER, M. MITCHENER AND A. P. NASON
and the diet with brown sugar. These
three groups comprised series A.
A second series of rats, series B, was
born April 4, 1968. One group of 40,
equally divided as to sex, was given the
torula diet with white sugar and the basal
water in order to repeat the observations
on series A. A second group of 37 was given
the torula diet with raw sugar and the
basal water.
Rats were weighed at 2-week intervals
until 3 months old and monthly thereafter.
After fasting 18 hours, serum cholesterol
was measured by the method of Huang et
al. (6) and serum glucose by that of
Washko and Rice ( 7 ), using premixed reaagents.10 The diets were analyzed for essen
tial trace metals by atomic absorption
spectrophotometry." The results are shown
in table 1.
The second type of diet was our regular
starch diet of 60% whole untreated Balbo
rye flour, 30% dried skimmed milk, 9%
* Pure lard shortening, Tobin Packing Co., Albany,
N Y
s The salt mixture contained: (%) CaCO:i 48, MgCOa
25 MgS04 8H,O 1.6, NaCl 6.9, KC1 11, KH2PO4 21,
FeSO4-7H2O 3, KI 0.008, Na2SeO4 O.OÇ1,
cystine 6. The
vitamin mix provided: (mg/100 g diet) choline chlo
ride 100, niacin 10, vitamin Bi2 (0.1% in mannitol)
0.1, d-Ca pantothenate 2, thiamin-HCl 0.4, riboflavin
0 25 pyridoxine 0.2, menadione 0.1, biotin 0.05, folie
acid 0.2, crystalline retinyl acetate 1.4 (4000 IU), crys
talline vitamin D»0.01 (400 IU).
6 Domino brand, American Sugar Co., New York.
7 See footnote 6.
8 Supplied by John L. Hickson, Ph.D., Sugar Re
search Foundation, New York, from a shipment of
Philippine sugar to American Sugar Co., Brooklyn,
N. Y. It contained 97.88% sucrose, 0.73% invert sugar,
and 0.7% undefined material.
9 Random bred pregnant females were obtained from
Blue Spruce Farms Inc., Altamont, N. Y.
">Berkeley Medical Instruments, N.E. X-Ray and
Electronics, 157 Sutherland Road, Brookline, Mass,
i' Model 303, Perkin-Elmer Corp., Norwalk, Conn.
TABLE 1
Essential trace metals in torula diets containing
Downloaded from jn.nutrition.org by on November 25, 2008
commerial lard," 4% sait mixture (pro
viding 0.24% cystine) and 1% vitamin
mixture 5 modified from that of Schwarz
and Mertz (3) was prepared. Sodium sele
nite was added to the salt mixture in an
amount to provide an additional 0.2 ppm
selenium in the diet. The vitamin E con
tent was not measured; commercial lard
is said to contain 2.0 mg/100 g (4). It is
estimated that the diet contained 3 ppm
vitamin E. No antioxidants were added,
except selenite, nor did the lard contain
any according to the label. Compared to
the diet of Schwarz and Mertz (3), our
diet differed in having: less sucrose (50%
vs. 59%), more fat (15% vs. 5%), the
use of commercial lard instead of vitamin
E-free lard, and added cystine. There were
minor differences in the salt and vitamin
mixes.
The diet was mixed twice weekly. Three
types of cane sugars were given ( 1) : re
fined white sugar,6 dark brown sugar 7 and
raw sugar." The basal drinking water con
tained (ppm): zinc 50, manganese 10,
copper 5, cobalt 1, molybdenum 1 as sol
uble salts (5) given to all animals. To one
group of rats, 5 ppm chromium as the
acetate was added to the water.
Long-Evans rats of the BLU: (LE) strain
were obtained as random bred pregnant
females.9 Litter mates born August 20,
1967 were divided into three groups of 20
of each sex, four to a cage. When weaned,
they were given the basal water and the
torula diet with white sugar, the basal
water with 5 ppm chromium added and the
diet with white sugar, or the basal water
various sugars, and standard diet '
sugar
sugar
sugar
MetalCrMnCoCuZnMoNi*Sr*White
dietß9/90.083.050.081.9528.60.090.085.43Brown
dietna/g0.141.55<
dietß9/90.1412.70.35-0.461.3622.30.250.2-0.75.23
starch
dietIÕ9/90.164.350.234.5827.70.040.3014.45Raw
0.031.1817.10.040.137.15Regular
1 Torula diet of 30% yeast, 50% sugar, 15% lard and 5% salt mixture. Regular diet of 30%
skimmed milk, 60% whole rye flour, 9% corn oil and 1% sodium chloride, with iron added. All
values on dry weight basis.
* Possibly essential.
CHOLESTEROL AND GLUCOSE OF RATS
with two groups fed the torula yeast diet
(series B). This group received 5 ppm
chromium in drinking water. More than
half were living in August, 1970.
RESULTS
Body weights. The weights of rats up
to 12 months of age given the torula yeast
diet and various sugars have been reported
(1). The growth rates of rats fed the reg
ular diet were similar to those of series B
fed the yeast diet up to 1 year of age. In
general, the torula diets were adequate for
growth up to a year of age, but weight
was lost at 18 and 24 months in all groups
but those fed the regular diet (P < 0.01
or 0.001). Losses in males were 25 to 44%
and in females, 11 to 28% . Unlike the case
of rats fed the regular diet (10, 11), chro
mium had no significant effect on growth
of rats fed the yeast diet.
Survival. Rats given the torula diet sur
vived poorly as compared with those given
the regular diet. At 18 months of age, over
all mortality of 103 rats of both sexes fed
the regular diet was 5.9% and at 24
months, it was 21.4%. Of 198 rats fed the
torula yeast diet, mortalities at these ages
were 24.2% and 60.1% respectively. Dif
ferences by chi-square analysis were sig
nificant at both ages (P < 0.0005). This
diet was therefore adequate for growth but
not for survival. Survival of the various
groups is shown in table 2. Median life
spans of rats fed brown sugar were greater
than those of the other groups, but less
than a comparable series given the regular
diet. Mortalities at 24 months of age were :
white sugar A 58% , white sugar plus chro
mium 75% , brown sugar 40% , raw sugar
59% , white sugar B, 72% . All but five rats
in series A (95.9% ) were dead by 30
months of age, whereas about 50% of rats
fed the regular starch diet have regularly
survived to 30 months of age (P < 0.0001 ).
Longevities, or mean ages of the last five
survivors, were greater in the brown-sugar
groups than in the white-sugar groups and
the male white-sugar-plus-chromium group;
they were also greater in the raw-sugar
groups than in the comparable white-sugar
groups of series B.
Serum cholesterol, torula yeast diet.
Fasting serum cholesterol levels of rats
given the torula yeast diet at 2 or 3 ages
Downloaded from jn.nutrition.org by on November 25, 2008
corn oil and 1% iodized sodium chloride,
with iron and vitamins added (5). Con
centrations of essential trace elements are
given in table 1. Groups of 50 or more rats
of each sex were fed this diet, differing
only in the drinking water supplied. For
pilot studies, smaller numbers were used.
There were five series. Series 1 control
animals were given the basal water with
no chromium. Other groups were given the
basal water with 5 ppm chromium and the
basal water with 5 ppm cadmium. In this
series, serum cholesterol levels were mea
sured by the method of Abel et al. (8); the
data have been reported (9) and are in
cluded here for completeness.
Series 2 control animals were given the
basal water with 1 ppm chromium added.
A group of females of the same age was
fed the basal water with 5 ppm chromium
added.
Series 3, 4, and 5 control animals were
fed the basal water plus 5 ppm chromium.
In series 3 was a full group given 5 ppm
nickel, a full group given 50 ppm cadmium,
and a pilot group of 24 rats given 10 ppm
molybdenum, in addition to the basal water
plus 5 ppm chromium. One group was
given 12 ppm chromium.
Because molybdenum is in the same
periodic subgroup as chromium, and there
fore might exert similar or opposite actions
biologically, a full complement of rats were
fed 50 ppm molybdenum as molybdate in
doubly deionized water without any other
elements but those in the diet, and com
pared to their litter mate controls in series
4. Another 20 rats of each sex were given
the basal water plus 50 ppm cadmium and
5 ppm chromium. The control rats were
given basal water plus 5 ppm chromium.
Series 5 consisted of a control group of
young rats given 5 ppm chromium and a
full complement of animals fed the basal
water without chromium. They were not
concurrent with each other.
Serum of each of 12 rats from each
group was analyzed for cholesterol and
glucose, usually at yearly intervals. All rats
were born and housed in a laboratory espe
cially constructed to avoid environmental
metallic contamination (7).
In order to compare growth rates, one
group of 103 rats given the regular starch
diet (series 4) was weaned simultaneously
249
250
H. A. SCHROEDER, M. MITCHENER AND A. P. NASON
TABLE 2
Survival of rats fed the torula yeast diet and various sugars
DietBrown
AWhite sugar
AWhitesugar
AWhitesugar + Cr
BRaw sugar
BStarch
sugar
«Brown
diet, series 3
±22.8864
±9.5»732
>•«801
±12.5
±8.3849
±13.6«11
±27.8921
60
2592680872782667636
±14.2823
±11.2»917±31.75759
*63069891275%
4870±21.
±27.4«1304
±36.0
1Mean ±SEMof last five surviving rats.
2Differs from brown sugar group by Student's t, P < 0.005, or by chi-square P < 0.005.
»
P < 0.001.
«
Differs from white sugar group, P < 0.005, of respective series.
*P < 0.025.
8Series 3 rats were started July, 1966 and were not concurrent with series A, started August, 1967 or series B,
started April, 1968. Series 4 rats were littermates of series B, but more than half were alive in August, 1970:
thus, series is incomplete. Series 3 is shown for comparison.
TABLE 3
Fasting serum cholesterol levels ' in rats fed various sugars and a torula yeast-lard-sugar
SugarWhite
AWhite
rats1212121212412121212121212MalesCholesterol
levelmg/100
ml74
±2.7104
±6.8»109±4.168
diet
rats—129—11612121212121212FemalesCholesterol
levelmg/100
ml—110±11.4123
———<
±14.3—59
CrABrown
+
±3.157
±3.5«85
0.001<
±5.5»62
0.001<
AWhite
755±4.
0.025<
±3.383
0.001<
±3.4»60
0.001——<0.05<
BRawB«Agedays150324680151322715157318709128486128486No.
±3.287±3.1»53
±1.762
±3.5«P2NS< 0.001No.
±4.481
0.001<
±8.3»69
0.025—<
±3.177
=93±2.2
0.001<
±2.9»60
0.025—<
±1.596±2.3344
±1.776
±2.3»pz.
0.001<
0.001
1Mean ±SEM.
2 Significance of difference from white A or white B by Student's t at comparable age.
3Value differs from that of proceeding age, P < 0.001.
«
P < 0.025.
»
P < 0.005.
6The data up to 324 days of age for series A and at 128 days of age for series B have been reported (1),
and are included here for comparison.
are shown in table 3. In series A, the levels
of male rats fed white sugar rose from 5
to 11 months of age, and remained rela
tively unchanged in both sexes from 11 to
23 months of age. When chromium was
given, levels were significantly lower at
both older ages. Cholesterol levels of rats
fed brown sugar were significantly lower
at all ages than were those of the whitesugar groups, and differed little from those
fed white sugar plus chromium. In series
B, cholesterol levels of rats fed raw sugar
were significantly lower at 4 and 16 months
than were those of rats fed white sugar.
Downloaded from jn.nutrition.org by on November 25, 2008
AWhite sugar
AWhitesugar
AWhitesugar + Cr
BRaw sugar
BStarch
sugar
diet, series 3 eNo.
i924
rats202019191752212021212052Median
age800740601 dead8878206867667949748417147397087721050daysMales9038607428108481057Females9318
dead
90%
survivor101089075683188812329608609378479251347Longe
CHOLESTEROL AND GLUCOSE OF RATS
mium, levels were depressed at 16 but not
at 30 months of age.
In series 4, 50 ppm cadmium did not
result in significant differences from con
trol cholesterol levels of older rats. These
results are in conflict with those in series
3. The feeding of 50 ppm molybdenum
alone was associated with lower levels of
serum cholesterol in older male rats than
in their control counterparts at two ages,
and lower levels in females at one age.
The two groups of series 5 were of dif
ferent ages, the group without chromium
being 19 months older than that fed 5 ppm
chromium. Comparison at the youngest
ages showed lower cholesterol levels in the
latter than in the former.
Although not strictly comparable, as the
groups were not concurrent, table 4 indi
cates that when 5 ppm chromium was fed
to rats, fasting serum cholesterol levels
were usually lower (in five of six pairs of
analyses in males comparable as to age
and in four of seven pairs in females)
than when chromium was not given. The
effect of cadmium in the various series
was inconsistent. The feeding of molyb
denum at 10 ppm plus chromium at 5 ppm
was associated with higher cholesterol
levels in young rats than were 50 ppm
molybdenum alone; levels in older rats
were similar.
Age-linked changes in serum cholesterol
levels occurred in the various series but
were not consistent. In males fed 5 ppm
chromium, levels were lower in older than
in younger rats at one or two intervals in
series 3, and were higher in series 4. In
females, they were lower in no series, and
higher in series 4. Lower values at older
ages also occurred in males fed nickel and
in the series 5 no-chromium groups. Higher
values at older than at younger ages were
found in series 1 and 3 cadmium-fed ani
mals, in females fed 12 ppm chromium,
and in series 4 animals fed 50 ppm molyb
denum. They were also higher in series 2
controls fed 1 ppm chromium at one inter
val. It is noteworthy, however, that fasting
serum cholesterol values greater than 100
mg'lOO ml occurred largely in rats fed
0 to 1 ppm chromium (11 of 19 analyses)
or cadmium (3 of 13 analyses), and were
unusual in rats fed 5 ppm chromium (2
Downloaded from jn.nutrition.org by on November 25, 2008
Age-linked changes were evident in four
of the five dietary groups. In rats of series
A, cholesterol levels rose significantly from
11 to 23 months of age, in males and fe
males fed white sugar plus chromium, by
28 and 22 mg/100 ml, respectively, and
in those fed brown sugar by 28 and 16
mg/100 ml, respectively, but not in those
fed white sugar only. In series B, from 4
to 16 months of age, levels rose in males
and females fed white sugar by 27 and 36
mg/100 ml, respectively, and in those fed
raw sugar by 9 and 32 mg/100 ml, respec
tively. From 5 to 11 months of age, de
clines in cholesterol levels occurred in
males fed white sugar plus chromium and
brown sugar, but not in those fed white
sugar only. A progressive rise in serum
cholesterol with age was found in females
fed brown sugar. Thus, chromium and
brown sugar did not inhibit the rise with
age in older male rats, as they had in
younger ones.
Serum cholesterol, starch diet. Fasting
serum cholesterol levels of rats given the
regular starch diet without chromium, with
three levels of chromium, and with nickel,
molybdenum, or cadmium are shown in
table 4. For comparison and completeness,
the first series, which has been reported
(9), is included. In series 1, both chromium
and cadmium were cholesterolytic in older
males but not in females. Cadmium-fed
chromium-deficient rats had increased cho
lesterol levels at the oldest age.
In series 2 males fed 1 ppm chromium
appeared partly to modify age-related
changes in serum cholesterol levels as
shown in control males of series 1. There
was little effect in females. When a group
of females was given 5 ppm chromium,
levels were lower than those of rats fed 1
ppm.
Series 3 rats fed 5 ppm chromium
showed good control of cholesterol levels.
The addition of nickel at 5 ppm was ac
companied by lower levels than the con
trols in males but not in females. The ad
dition of 50 ppm cadmium resulted in
significant elevations of cholesterol levels
in males as compared to the controls, with
one value higher in females. The addition
of 10 ppm molybdenum appeared to have
no effect on cholesterol levels in either sex.
When female rats were given 12 ppm chro
251
252
H. A. SCHROEDER, M. MITCHENER AND A. P. NASON
TABLE 4
Fasting serum cholesterol levels ' in rats fed chromium, nickel, molybdenum and cadmium,
according to age; starch diet
waterControl
Metal in basal
Males
Females
levelmg/100
levelmg/100
ml102
±4.5108
±4.4123
±8.291
±4.877
±6.693
±7.676
±4.768
2.889
±
±8.7'111±8.776
CrChromium
1, 0
ppmCadmium5
ppmControl 5
ppmControl
1
±7.395±11.2—101
0.001<
±8.2114
0.05NS_NSNS————<
0.01<
±9.5—87±9.8113±9.0«72
0.001<
0.001<0.01—————————<0.01<
±5.3120
5116±6.1109
±9.3
±2.9'78
±2.192
±5.3»—86
±4.063±4.172
0.001_—_<
5
3Chromium
ppmChromium
ppmNickel
±3.267±5.1»——75
5
±5.277
±5.262
±2.786
±3.4»75
12
+chromium
5 ppm
ppmMolybdenum5
±2.649±4.8579
0.05NSNSNS—NSNS<
±3.880±4.777
0.01—NS—<
ppm+
10
chromiumCadmium
5 ppm
±2.483
±5.464
±4.276
±5.874
ppm4- 50
chromiumControl
5 ppm
Downloaded from jn.nutrition.org by on November 25, 2008
2Chromium
ppmChromium
ml—80
±4.2107
±2.488
5111±6.1
±2.8*84
0.005<
0.05————
0.001———Agedays—510761—510761—51076040565769871870940572648091234266315
±5.960
±12.446±1.788±2.45100
4Chromium
±3.494±3.7585
5 ppmAgedays360510761360510761360510760405657668718—402718——342663151315231576917129476810Cholesterol
±7.3Pi———NS<
Molybdenum
50 ppm
(in doubly
deionized water)
135
477
813
Cadmium
50 ppm +
5 ppm chromium
719
Control 5, 5 ppm
chromiumNo
chromium
'9011522857184
NS
±5.8PZ_—__—<
51 ±2.2
76±4.65
76 ±3.5
< 0.025
< 0.005
135
477
813
57±1.5
74 ±5.4 '
97 ±4.2 5
99 ±5.2
NS
719
97 ±4.3
±1.9114
±5.0123
±10.178
±4.8*<
NS
< 0.005
^0.05
NS
0.005<
±4.0110±2.791 0.001<
0.005—~~9012722857174
0.001——
±6.186
±6.0<
1 Mean ±SEM.
2 P is significance of difference of value from that of control group of comparable or nearest age by Student's
t. NS: Not significant.
3 Value differs from that of preceding age, P < 0.025.
< P < 0.05.
5 P < 0.001.
«P < 0.005.
7 Analyses were made on 12 rats in each group. Series 1 has been reported (9). All rats were given basal water
except group fed molybdenum
50 ppm. Basal water contained 50 ppm zinc, 10 ppm manganese,
5 ppm cop
per, 1 ppm cobalt, 1 ppm molybdenum.
Series 1 was started Feb. 1961; series 2, Sept. 1963; series 3, July 1965;
and series 4, April, 1968; therefore each metal-fed group in each series was run concurrently.
The series
5, 5 ppm chromium group was begun May 1970, and the 0 ppm chromium group was started January 1969;
therefore these two treatments were not concurrent.
253
CHOLESTEROL AND GLUCOSE OF RATS
and 102 ±4.5. Likewise, males fed 5 ppm
chromium had fasting glucose levels at 14
and 24 months of age of (mg/100 ml)
101 ±5.5 and 83 ±2.9; when fed 1 ppm
chromium they had 117 ±2.2 and 107
±3.5, and when given no chromium they
had 112 ±3.7 and 134 ±5.1, respectively.
Nickel was present in sizable amounts
in brown sugar (table 1). Therefore, pos
sible glycolytic effects of nickel in rats fed
the starch diet were evaluated in series 3
animals. No depression of fasting serum
glucose occurred in males at two intervals,
nor in older females, but at 12 and 13
months of age, levels were lower ( 78 ±4.2
mg/100 ml) in the nickel-fed females than
in the controls (90 ±4.9 mg/100 ml, P
< 0.05).
In series 3 rats given 10 ppm molyb
denum plus 5 ppm chromium, there ap
peared to be a synergistic action of the
two trace elements on fasting glucose
levels. At the one comparable interval,
males and females had serum glucose
levels of 78 ±5.1 and 63 ±1.7 mg/100
ml, differing significantly from their con
trols of 101 ±5.5 and 90 ±2.1 mg/100
ml, respectively (P < 0.005 and P < 0.001).
This phenomenon did not appear in rats
fed 50 ppm molybdenum without chro
mium. In fact, male levels did not differ
TABLE 5
Fasting serum glucose levels ' in rats fed various sugars
SugarWhite
levelmg/100
ml107±4.5104±4.0109
levelmg/100
ml—118±3.4107
AWhitesugar
+CrBrown
sugar
±3.197
±2.9109
3141±3.7
«86±1.786
±5.6
±7.6—83
0.05NS<
±4.6134
0.001<
498±4.6
0.005—NSNS——<
0.001<
±3.085±4.13126
0.001<
sugarWhite
±3.8105
0.001<
7.5«122
:±
O.O01——< 4132it3. 7
±3.888±8.14119±5.194
±3.8105
BRaw sugar
±5.5«110±4.9105
sugar 5Agedays150324680151322679157318686128457128456MalesGlucose
0.025NS
0.05NSGlucose
±3.3«P2————<
±5.5FemalesP2———<
1Mean ±SEM.
2 Significance of difference between value shown and that of preceding group of comparable age,
by Student's t. NS: Not significant.
3Value differs from that of preceding age in same group, P < 0.01.
4 Value differs from that of preceding age in same group, P < 0.001.
oThe data up to 324 days of age for series 1 and at 128 days of age for series 2 have been reported
(1) and are included here for comparison. There were 12 rats in each group, except as shown in
table 3.
Downloaded from jn.nutrition.org by on November 25, 2008
of 36 analyses). This distribution is sig
nificant by chi-square analysis (P < 0.001).
Fasting serum glucose, torula yeast diet.
Rats of series A fed white sugar plus chro
mium or brown sugar exhibited signifi
cantly elevated fasting serum glucose levels
at about 23 months of age compared to
those at 11 months of age (table 5), al
though levels of males fed brown sugar
were lower at all ages than those given
chromium. They were little changed in the
group given white sugar alone. In series B,
there were age-linked declines in fasting
glucose levels of rats of both sexes fed
white sugar and in females fed raw sugar.
Fasting serum glucose, starch diet.
Fasting serum glucose levels were obtained
in 68 groups of rats fed the starch diet
at the times cholesterol levels were as
certained. The hypoglycémie effect of
chromium has been reported (11) and sub
sequent studies have confirmed this phe
nomenon. For example, female rats fed 12
ppm chromium in water for 16 months had
glucose levels (mg/100 ml) of 75.3, com
pared to 104 ±4.9 in rats given 5 ppm
chromium, 114 ±5.4 in rats given 1 ppm
chromium and 138 ±4.8 in rats given no
chromium at comparable ages. At 24 to 30
months of age, values were (mg/100 ml),
respectively: 87 ±3.8, 85 ±4.6, 114 ±5.8
254
H. A. SCHROEDER, M. MITCHENER AND A. P. NASON
tively, smaller by 27.6% and 19.4% than
those of rats fed the starch diets (12).
There were no consistent
differences
among the 10 groups in the ratio of heart
size to body weight at death. There were
also no obvious differences between the
groups in time of onset of the paralysis,
in tumors, or in blanching of the incisor
teeth. Weight loss at death compared to
weight at 18 months of age was less in
females fed brown sugar (23.5% ) and
both sexes fed raw sugar (males 28.2%,
females 23.2% ) than in the groups given
white sugar with or without chromium
(males 35.8 to 39.2%, females 33.2 to
41.4% ).
DISCUSSION
A factor responsible for a low or normal
serum cholesterol in rats is chromium (III).
Therefore, one sign of chromium deficiency
is a relatively elevated serum cholesterol
level. This essential trace metal was active
in rats fed both our regular starch diet and
the torula yeast diet, and in hypercholesteremic rats fed cholesterol (2). Compared
to chromium-deficient male rats given the
regular starch diet (0.14 /«g/gchromium),
the addition of 1 ppm chromium in drink
ing water was associated with significantly
lower serum cholesterol levels and the ad
dition of 5 ppm resulted in further lower
ing. In females 5 ppm were necessary to
produce relatively low levels and 12 ppm
were more effective at 16 months of age
and equally effective at 30 months of age.
None of 12 other elements showed this
type of action (13), although elements of
the first and second transitional groups fed
to rats tended to be associated with lower
cholesterol values than the heavier metals
and the nonmetals.
When serum cholesterol levels of older
rats fed 14 trace elements were evaluated
according to the positions of the elements
in the periodic table, interesting trends ap
peared (13). Elements of periods 4 and 5
were assigned arbitrary numbers from 2
to 27 according to relative locations in the
two periods and arranged in the order:
zirconium, vanadium, niobium, chromium,
molybdenum, nickel, cadmium, germani
um, tin, lead, arsenic, antimony, selenium,
tellurium. Cholesterol levels of rats fed
each of these elements showed positive
Downloaded from jn.nutrition.org by on November 25, 2008
from their controls at 4 months of age,
were significantly higher than their con
trols at 16 months of age and were lower
at 27 months of age, whereas female
levels were significantly lower than their
controls at 4 months of age, were high at
16 months of age and did not differ at 27
months of age. Thus, the effects were in
consistent. A comparison of serum glucose
levels of rats of both sexes fed 5 ppm chro
mium and 10 ppm molybdenum with levels
of rats fed 50 ppm molybdenum without
chromium showed significant differences
(P < 0.001) at 4 to 5 months and 11 to
16 months of age, respective ranges being
63 ±1.7 to 78 ±5.1 and 90 ±2.3 to 128
±4.0 mg/100 ml.
Fasting glucose levels of rats fed 5 ppm
chromium, compared to those fed 50 ppm
cadmium in addition to 5 ppm chromium,
did not differ from their controls at two
intervals and were higher than their con
trols at the oldest interval.
Effect of torula diet in older rats. Rats
fed the torula yeast diet for 18 months
began to exhibit nervous irritability and
hyperactivity of the lower extremities with
twitching and spasms. Paralysis of the hind
quarters slowly developed during the fol
lowing 3 months, until it became total.
Weight loss was evident at 2 years of age,
which progressed to the point that at death,
males had lost 37.5% of their weight at
18 months of age and females had lost
30.1 % . Scabby skin lesions were frequently
found, but loss of hair occurred in only
5% of the animals. Cannibalism was found
in 21.2%, and in several instances ani
mals had eaten their own tails, suggesting
that the paralysis was sensory as well as
motor. Permanent lateral or dorsal spinal
curvatures were common, and dental caries
were often seen. There was a remarkable
loss of subcutaneous and abdominal fat;
it could not be found in 67.2% of cases.
Of all rats autopsied, paralysis occurred
in 90.4% of series A and 84.6% of series
B.
The loss of weight was somewhat as
sociated with reduction in the size of the
heart relative to that of the body. Ratios
of heart weight to body weight X 1000
varied from 3.79 to 5.92 in males and
4.10 to 5.55 in females, the mean heart
sizes being 1084 mg and 765 mg, respec
CHOLESTEROL AND GLUCOSE OF RATS
255
Downloaded from jn.nutrition.org by on November 25, 2008
correlations (r) with relative position in nal considering the rise in older animals,
males of 0.53 (P < 0.025) and in females
whereas male rats fed the starch diet re
of 0.66 (P < 0.01). Thus, the first six ele
quire at least 1.0 ^g and females require
ments were more likely to be associated
1.5 fig/100 g body weight/day. According
with lower cholesterol levels than the last to these calculations, commercial diets
six, males showing mean values of 83 ±3.2 which have 1 to 1.2 Mg/g chromium (16)
and 97 ±4.8 mg/100 ml (P < 0.025), and would supply 6 to 7.2 fig/100 g per day, a
females 78 ±4.8 and 101 ±3.3 mg/100
luxus amount. It is possible that the
ml (P < 0.001), respectively.
amount of chromium actually absorbed
According to these data, molybdate in from the starch diet is less than that from
large amounts (50 ppm) can exhibit a the torula diet, and that optimal amounts
chromium-like action on serum cholesterol
for rats are 2 to 3 /¿g/100g per day.
of rats, but not on fasting serum glucose
If the requirement of chromium for cho
levels. Glycolytic effects occurred only lesterol homeostasis in the rat can be
when chromium was fed in addition. Such applied to man, a 70-kg man would need
actions of molybdenum probably depend
at least 10 fig/kg body weight per day, or
upon the similarity of its atomic structure
700 ftg/day. Analysis of hospital diets has
to that of chromium.
shown
102 ^g/day (16), and others have
Chromium (III) salts are poorly absorbed
from the gastrointestinal tract, 0.1 to 1.2% found 52 to 115 /ig/day in this country
(mean 0.5 ±0.3% ) appearing in the urine and 120 to 172 /ig/day in foreign coun
tries, by atomic absorption spectrophotomof patients (14). In rats, 2 to 3% is ab
etry
(15). Tipton et al. (17, 18) using
sorbed, but according to Mertz (15), nat
urally occurring complexes are easily ab emission spectrography, recorded average
chromium contents of long-term diets of
sorbed. It is probable that the chromium
in sugar is in a readily absorbable form. five persons at 200 to 400 ^g/day. If chro
If this is so, and all of the chromium in mium is concerned with cholesterol homeo
the torula diets were absorbed, at a daily stasis in man, it is clear that many Amer
ican diets are inadequate in chromium and
ingestion of 6 g food/100 g body weight,
rats would take into their bodies 0.96 /ig a rise in serum cholesterol with age could
from the brown sugar diet, 0.84 /ig from be expected.
The apparent hyperglycémie effect of
the raw sugar diet and 0.48 ¡igfrom the
white sugar diet per 100 g weight of added chromium or brown sugar with age
rat/day. Rats on the regular starch diet was surprising. In young rats, however,
drink about 7 ml water/100 g body weight we were unable to demonstrate a hypo
per day, or 7 and 35 /tg chromium added glycémieeffect of chromium when the
at 1 or 5 ppm. At an absorptive rate of torula yeast diet was used (11), although
2 to 3%, they would take in 0.14 to 0.21 it occurred regularly in rats given the rye—
^.g when given 1 ppm chromium in water milk-corn oil diet (11, 19). This difference
and 0.70 to 1.05 /¿gwhen given 5 ppm, in the action of chromium in sucrose and
starch diets deserves examination.
in addition to the amount in the diet.
The torula yeast diet was adequate for
On these bases and assumptions, the
eventual growth compared to our starch
following are the approximate calculated
amounts of chromium absorbed by rats in (rye) diet, although rats grew slowly for
the various groups: (¿ig/100g body weight 2 months, a characteristic of chromium
per day) white sugar 0.5; brown sugar 1.0; deficiency (10, 11). It was inadequate for
maintenance of body weight after 1 year
raw sugar 0.8; white sugar plus chromium
of age and was inadequate for long sur
1.2 to 1.5, starch diet with 1 ppm chro
vival. The diet was deficient in tocopherol,
mium 1.1, starch diet with 5 ppm chro
mium 1.5 to 1.9, starch diet with 12 ppm in spite of the addition of selenium, which
chromium 2.5 to 3.5. Therefore, in terms can lessen requirements for vitamin E.
of cholesterol homeostasis, rats given the Pentschew and Schwarz (20) have de
torula diet require at least 0.8 to 1.0 ¿ig scribed symptoms similar to those exhibited
absorbable chromium/100 g body weight by our rats, when a torula yeast diet de
ficient in vitamin E but supplemented with
per day, an amount which may be margi
256
H. A. SCHROEDER, M. MITCHENER AND A. P. NASON
have reflected the soils on which the sugar
cane was grown.
These experiments in rats point to an
intimate relationship of chromium and cho
lesterol metabolism. Therefore they sug
gest that human cholesterol levels may be
elevated partly as a result of excess con
sumption of refined white sugar, which in
turn is responsible for the chromium de
ficiency found in tissues of United States
subjects but not in those from foreign
countries (16, 23).
ACKNOWLEDGMENTS
We are indebted to Miss Lois A. Bradstreet and Steven Skibniowsky for the
analyses of serum and to Dwight Bell for
assistance in caring for the animals.
LITERATURE CITED
1. Schroeder, H. A. 1969 Serum cholesterol
and glucose levels in rats fed refined and less
refined sugars and chromium. J. Nutr. 97:
237.
2. Staub, H. W., G. Reussner and R. Thiessen,
Jr. 1969 Serum cholesterol reduction by
chromium in hypercholesteremic rats. Science
166: 746.
3. Schwarz, K., and W. Mertz 1959 Terminal
phase of dietary liver necrosis in the rat
(hepatic hypoglycemia). Metabolism 8: 79.
4. Diem, K. (ed.)
1962 Scientific Tables, ed.
6. Geigy Pharmaceuticals,
Ardsley, N. Y.
5. Schroeder, H. A., W. H. Vinton, Jr. and J. J.
Balassa
1963 Effect of chromium, cadmi
um and other trace metals on the growth
and survival of mice. J. Nutr. 80: 39.
6. Huang, R. C., C. P. Chen, V. Wefler and A.
Raferty
1961 A stable reagent for the
Liebermann-Burchard
reaction. Clin. Chem.
7- 542
7. Washko, M. E., and E. W. Rice 1961 De
termination of glucose by an improved enzy
matic procedure. Clin. Chem. 7: 542.
8. Abell, L. L., B. B. Levy, B. B. Brodie and F.
E. Kendall 1952 A simplified method for
the estimation of total cholesterol in serum
and demonstration of its specificity. J. Biol.
Chem. 195: 357.
9. Schroeder, H. A., and J. J. Balassa
1965
Influence of chromium, cadmium and lead
on rat aortic lipids and circulating choles
terol. Amer. J. Physiol. 209: 433.
10. Schroeder, H. A., W. H. Vinton, Jr. and J. J.
Balassa
1963 Effect of chromium, cadmi
um and lead on the growth and survival of
rats. J. Nutr. 80: 48.
12Our diet differed from that used by Pentschew and
Schwarz (20) in having less sugar (50 vs. 59% ) and
more lard (15 vs. 5% ). The salt mixtures were simi
lar, except that ours contained no fluoride, aluminum,
manganese or copper, the last two being supplied in
water. Vitamin mixtures were identical, except that
we added vitamin B,2.
Downloaded from jn.nutrition.org by on November 25, 2008
selenium was fed.12 If the commercial lard
contained 20 mg/kg vitamin E (4), adult
rats would have ingested in the diet 0.06
mg/day, an amount too small to prevent
deficiency at older ages but apparently
enough for growth.
Mertz and Roginski (21) have shown
that chromium-deficient rats severely de
ficient in vitamin E to the point that growth
was retarded and early deaths occurred,
failed to respond to chromium supplemen
tation, whereas rats given vitamin E re
sponded to chromium by increased growth.
In this laboratory, young rats severely de
ficient in both vitamin E and chromium
had lower fasting serum glucose levels than
did rats given chromium; this difference
also was found between chromium-supple
mented and chromium-deficient rats given
vitamin E (11). Mertz' rats, however, were
fed soy protein and given vanadium and
molybdenum in water, whereas our rats
were fed the torula yeast diet and given
no added vanadium. In spite of these dif
ferences, it is possible that vitamin E is
required for chromium to act on growth,
but not on serum glucose levels. We have
been unable to demonstrate hypoglycémie
activity of vanadium, zirconium, niobium,
antimony (12), germanium,
tin (22),
cadmium, selenium or tellurium in rats
fed the starch diet, although males fed
lead (12) or arsenic (22) had low fasting
glucose levels.
Dark brown sugar as used in these ex
periments is a semirefined sugar to which
molasses has been added. Molasses con
tains most of the trace and bulk elements
which were present in the juice of the cane.
The "raw" sugar used was a light tan color,
crystalline, freely flowing and had ob
viously been partly refined; it was purer
than the dark brown sugar.
From table 1 it is clear that the raw
sugar was somewhat refined in respect to
the essential trace elements manganese,
cobalt, copper and zinc. Analyses of the
three sugars for zinc, for example, showed
(¿ug/gdry weight) : white, 0.54; raw, 0.78;
brown 1.62. Ash weights were: white
0.16%, raw 1.28%, brown 2.24%. Chro
mium concentrations
were (/ig/g dry
weight) : white, 0.02-0.03; raw 0.06, brown
0.12-0.24. The small concentrations
of
manganese, cobalt and copper found may
CHOLESTEROL AND GLUCOSE OF RATS
18. Tipton, I. H., P. L. Stewart and P. G. Martin
1966 Trace elements in diets and excreta.
Health Phys. 32: 1683.
19. Mertz, W., E. E. Roginski and H. A. Schroeder
1965 Some aspects of glucose metabolism
of chromium-deficient rats raised in a strictly
controlled environment. J. Nutr. 86: 107.
20. Pentschew A., and K. Schwarz
1962 Sys
temic axional dystrophy in vitamin E defi
ciency in adult rats with implications in
human neuropathology. Acta Neuropathol. 1:
313.
21. Mertz, W., and E. E. Roginski
1969 Effects
of chromium
(III)
supplementation
on
growth and survival under stress in rats fed
low protein diets. J. Nutr. 97: 531.
22. Schroeder, H. A., M. Kanisawa, D. V. Frost
and M. Mitchener
1968 Germanium, tin
and arsenic in rats: effects on growth, sur
vival, pathological lesions and life span. J.
Nutr. 96: 37.
23. Schroeder, H. A. 1968 The role of chro
mium in mammalian
nutrition.
Amer. J.
Clin. Nutr. 21: 230.
Downloaded from jn.nutrition.org by on November 25, 2008
11. Schroeder, H. A. 1966 Chromium deficiency
in rats: A syndrome simulating diabetes
mellitus with retarded growth. J. Nutr. 88.
439.
12. Schroeder, H. A., M. Mitchener and A. P.
Nason
1970 Zirconium, niobium, antimo
ny, vanadium and lead in rats: Life term
studies. J. Nutr. JOO: 59.
13. Schroeder, H. A. 1968 Serum cholesterol
levels in rats fed thirteen trace elements. J.
Nutr. 94: 475.
14. Donaldson, R. M., Jr. and R. F. Berreras
1966 Intestinal absorption of trace quanti
ties of chromium. J. Lab. Clin. Med. 68; 484.
15. Mertz, W. 1969 Chromium occurrence and
function in biological systems. Physiol. Rev.
49: 163.
16. Schroeder, H. A., A. P. Nason and I. H. Tipton 1970 Chromium deficiency as a factor
In atherosclerosis. J. Chronic Dis. 23: 123.
17. Tipton, I. H., P. L. Stewart and J. Dickson
1969 Patterns of elemental excretion in
long-term balance studies. Health Phys. 16:
455.
257