See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/16108074

Biochemical composition of muscle in normal and semistarved human subjects:

Relevance to anthropometric measurements

Article  in  American Journal of Clinical Nutrition · August 1982

DOI: 10.1093/ajcn/36.1.131 · Source: PubMed

CITATIONS READS

60 300

6 authors, including:

Steven B Heymsfield Callum McManus

Pennington Biomedical Research Center Teesside University
1,044 PUBLICATIONS   76,978 CITATIONS    7 PUBLICATIONS   1,160 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

DXA bone and body composition View project

Metabolic Syndrome View project

All content following this page was uploaded by Steven B Heymsfield on 18 August 2014.

The user has requested enhancement of the downloaded file.

Biochemical composition of muscle in normal
and semistarved human subjects: relevance to
anthropometric measurements13
Steven B. HeymsfIeld,4 M.D., Victoria Stevens, B. S., Robert Noel, B. S.,

C4fford McManus, B.S., Janet Smith, M.M.Sc., R.D., and Daniel Nixon, M.D.

ABSTRACT Anthropometric methods aimed at assessing muscle size in undernourished

subjects assume a constant proportionality between the mass (i.e., size) and composition (specifically
protein-energy content) of this tissue. This assumption was examined in three autopsy groups:
controls I I, sudden traumatic
(n = death), early semistarvation (n = 6, acute preterminal disease),
and chronic sernistarvation (n = 34, severe weight loss over time). Results of semistarved groups
were expressed relative to respective control value. Early semistarvation produced no detectable

Downloaded from www.ajcn.org by guest on July 15, 2011

change in muscle mass, protein, or total energy content (per gram wet weight), although RNA and
glycogen were -50 to -70% of control value (p < 0.05). Chronic semistarvation caused muscle
atrophy (-54.2%), but not all measured constituents were reduced to the same degree. The results
were H,O-52.9%, collagen-46%, noncollagen proteins-65.3%, total lipids-40%, DNA-54.l%,
RNA-8l.7%, glycogen-90.3%, and total energy-59.6%. Muscle per unit mass in chronic
semistarvation thus reflects relatively more H,O and less protein and energy when compared to
normal tissue. About 85 to 95% of muscle protein-energy loss can be detected by anthropometric
measurements of muscle size; the remaining 5 to 15% depletion of protein and energy is masked by
muscle compositional changes. Proper interpretation of anthropometric data requires an under-
standing of these unmeasured but important compositional differences in normal and semistarved
muscle. Am J Clin Nutr 1982;36:l3I-l42.

KEY WORDS Arm muscle area, anthropometry, protein-energy malnutrition, semistarvation

Introduction of muscle bears an unknown but constant

relationship to muscle composition, specifi-
Muscle tissue comprises 40% of body cally muscle protein and energy content.
weight in healthy adult man, for a total mass Two lines of evidence in man, histological
of about 28 to 30 kg (1). Muscle is primarily and radiographic, suggest that muscle com-
water, but protein, which constitutes most of position is altered in PEM. Montgomery (5),
the nonaqueous portion of muscle, has been comparing microscopic sections of normal to
an important focus of the nutritionist. In PEM muscle, found increased connective tis-
protein-energy starvation, muscle protein un- sue, an enlarged extracellular space, and de-
dergoes catabolism and oxidation, a process creased myofiber mass (5). Radiographic
that supplies required metabolic fuel. This studies of muscle in PEM patients at our
potential calorie reserve has, over the years, center detected a decrease in x-ray muscle
been used to judge the overall depletion of
lean body mass (2) and the effectiveness of
I From the Departments of Medicine, Clinical Re-
protein-energy malnutrition (PEM) treat-
search Facility, Emory University School of Medicine,
ment (3). The accessibility of skeletal muscle Atlanta, GA 30322.
to anthropometric measuring techniques 2 Supported by NIH Contact 1-8565 1 and USPHS
makes muscle unique among the components Grant RR00039.
3 Reprints not available.
of lean body mass. However, the use in PEM
4 Teaching and Research Scholar of the American
of anatomic muscle mass indicators (e.g., arm
College of Physicians.
muscle area) (4) is based on an unproven Received August 25, 1981.
assumption: that in health and PEM the size Accepted for publication December 15, 1981.

The American Journal of Clinical Nutrition 36: JULY 1982, pp. 131-142, Printed in U.S.A. 131
© 1982 American Society for Clinical Nutrition
132 HEYMSFIELD ET AL.

density, which we hypothesized was a relative were patients not falling into one of the three categories
increase in muscle water or fat (6). defined below. At the beginning of the autopsy, anthro-
pometric measurements of the mid-upper arm were
Since the a priori assumption of anthro- made. The muscle specimens were then quickly excised,
pometric muscle mass measurement is that extraneous fat trimmed, and the tissue inserted into a
normal and undernourished muscle have polyethylene container before freezing at -80#{176}C.Bio-
similar biochemical composition per unit of chemical composition was then analyzed as described
below in iiBiochemical methods.”
tissue, but the above cited observations ap-
pear to contradict this notion, we investigated Subjects
the biochemical composition of normal and Of the 70 autopsies conducted during the study pe-
undernourished human muscle. Because of nod, 15 were excluded because more than 6h elapsed
the variety of diseases included in our subject between death and postmortem examination. Of the
pool, and the likelihood that other nutritional remaining 55 cases, 5 1 were selected for study who met
the following criteria (Table 1).
deficiencies compounded inadequate protein-
The first group of cases were those subjects without
energy intake, we henceforth refer to under- a history ofprior illness, who were 90 to 115% ideal body
nourished study groups by the more general weight (%IBW) (7), and died within several seconds or
term, semistarvation. minutes after traumatic injury.
The second group was composed of patients who had
no history of undernutrition before hospitalization, were
Methods

Downloaded from www.ajcn.org by guest on July 15, 2011

at or above IBW, and who were admitted for acute
illnesses which terminated in death within 3 days of
Design of experiment
admission. Dextrose (5%) and water provided the only
Anthropometry measures muscle size, or in essence, defmable source of nutritional support during this pe-
muscle mass. The experiment was designed to answer nod.
the question; is the loss in muscle mass in semistarvation The third group of patients all had a history of
accompanied by a proportionate reduction in muscle
total protein, nonstructural protein [the noncollagenous
protein fraction (NCP)], and energy content? This ques- TABLE 1
tion was examined by evaluating muscle mass and com- Patient diagnoses
position, and then applying the following equations. The Group
total mass of a given muscle is described by the equation:
Control Semistarvation
muscle mass = (H,O + DW), and (1)
Chronic
DW - (collagen + NCP + TL Early
Cancer Noncancer
+ DNA + RNA + glycogen), (2)
Causes of death
where DW is dry weight, and TL is total lipids. Total
Gunshot wound 7
muscle protein (TP) is then equal to the sum of NCP
Auto accident 4 1
and collagenous protein. Minerals account for about 1%
Malignant hyperten- 4 1
ofwet muscle weight and were excluded in our simplified
sion and CVAm
analysis. Equations (1) and (2) can be combined to
Acute myocardial in- 1
describe the major components of muscle mass or size:
farction
muscle mass (H,O + collagen Malignancy
+ NCP + TL + DNA + RNA + glycogen). (3) Melanoma 2
Pancreas 2
The energy content of wet muscle is equal to: Lung 4
Pineal 1
muscle energy (kcal/g wet weight)
Leukemia 1
= (K1TP) + (K,TL) + (K, glycogen) (4)
Ovary 2
where K = 5.65 kcal/g, K, = 9.4 kcal/g, and K., 4.1 Oropharynx I
kcal/g. Not all protein energy is available for metabolic Colon 1
processes, and utilizable protein energy can be calculated Ruptured aneurism 2
by subtracting 1.25 kcal/g protein to correct for urinary Congestive heart fail- 3
urea losses; K1 then become 4.4 kcal/g. ure 4
Our study was aimed at estimating muscle mass from Chronic lung disease I
anthopometric measurements, and the biochemical com- Renal failure
position (equation 3) and energy content of muscle Liver disease
(equation 4) by chemical analysis. These goals were Hepatitis 3
accomplished by harvesting the psoas and biceps muscles Cirrhosis 6
(50 g each) from autopsy subjects at Grady Memorial
Total 11 6 14 20
Hospital from January 1979 to June 1980. Autopsies
performed more than 6h postmortem were excluded, as a Cerebrovascular accident.
 MUSCLE COMPOSITION IN SEMISTARVATION 133

prehospitalization weight loss due to chronic disease. TP, NCP, and collagen. A 5-mi aliquot of the homog-
Death usually occurred after weeks of hospitalization enate was mixed with 5 ml of 0.1 N NaOH and used for
and additional weight loss (> 2 kg). Nutrient intake (as determination of TP, NCP, and collagen. NCP consists
judged from hospital records) during this period in all of primarily of intracellular proteins (sarcroplasmic, con-
these patients was far below the Recommended Dietary tractile, and mitochondrial) (12), whereas the collagen
Allowance for protein, energy, and other essential nutri- fraction is for the most part extracellular proteins (col-
ents (8). The third group was further subdivided into two lagen and elastin). For TP, 1 ml of the mixture was
types of subjects, those with and without cancer. This added to 9 ml of0.2 N NaOH. The remaining 5 ml were
latter subdivision was based on earlier reports of muscle centrifuged at 5000 x g for 30 mm. The supennatant
compositional changes unique to cancer patients (9- 1 1). containing NCP was collected and diluted 10-fold with
We subsequently refer to these three respective subject 0.2 N NaOH. The pellet was suspended in 9 ml of 0.2 N
categories as control group and early and chronically NaOH by a polytron homogenizer, and used for analysis
semistarved patients. It should be noted that the latter of collagen. Total protein, NCP, and collagen were de-
groups may also include patients with near total starva- termined by the heated Biuret-Folin procedure of Donsey
tion in the immediate hours or days before death. et al. (13).
TL and triglyceride. The Folch (14, 15) method of
Clinical methods purification was used to extract TL and its major sub-
Weight history was obtained from family and medical component, triglyceride, from 2 ml oftissue homogenate.
records on each subject. Body weight was measured at After the fmal washing in the procedure, the samples
autopsy, and %IBW calculated from Metropolitan Life were evaporated under nitrogen at 100#{176}C.The lipid was
Insurance tables (7). Postmortem anthropometric esti- resuspended in 2 ml of 2: 1 chloroform methanol solvent.

Downloaded from www.ajcn.org by guest on July 15, 2011

mates of mid-arm muscle area were made by measuring Seventy-five microliters were used for the determination
triceps skinfold thickness (TSF, in cm) with a Lange ofTG by the method ofSardesai and Manning (15), and
caliper, and mid-arm circumference (MAC, in cm) with 1 ml was evaporated in a heating block at I 10#{176}C using
a tape measure. The following equations (4) were then prewashed-weighed 13 x 100 tubes. After the drying
applied: procedure, the tubes were placed in a 200#{176}Cdrying oven
for 1 h to remove any residual chloroform. TL was then
arm muscle area (cm’)
determined by weighing these tubes.
(MAC - IT X TSF)’ _ 10, for men, and Glycogen, RNA, and DNA. A 10-ml aliquot of the
(5a) homogenate was used for sequential isolation and anal-
= 4r
ysis of glycogen, RNA, and DNA according to the
(MAC - x TSF)’ _ 6.5 5b modified procedure of Shibko et al. (16). The aliquot
for women = 4ir
was mixed with 2 ml of 70% perchlonic acid (PCA) and
allowed to incubate for 15 mm at 0#{176}C. The supernatant
muscle mass (kg) = height (in cm) [0.0 125
fluid containing glycogen was then removed, and the
+ (0.0034 x arm muscle area)]. (6) pellet washed with 5% PCA (1 X 10 ml). After a second
20 mm of centnifugation at 4000 x g, the supernatants
Muscle mass calculated from equation 6 should be were pooled and assayed for glycogen according to the
regarded as a rough approximation of total muscle mass procedure ofCarroll et al. (17). The pellet was incubated
(4); inclusion in the current report was for the purpose of with frequent vortexing in 10 ml of0.3 N NaOH at 37#{176}C
estimating how changes in muscle composition per gram for 2 h. The mixture was then treated with 2 ml of 70%
of tissue influenced overall body composition. If mea- PCA, placed in an ice bath for 10 mm, and then centri-
sured within 3 days of death, serum albumin was also fugated at 4000 x g for 20 mm. The supernatant con-
recorded. The control group had serum albumin levels tianing RNA was then separated, and the residue was
measured on postmortem intracardiac blood samples. washed with 5% PCA (1 X S ml); the pooled supernatants
were than analyzed for RNA (18).
Biochemical methods
The pellet obtained in the above step was incubated
We measured the mass of H,O and the six biochem- in 8 ml of 1.5% PCA for 30 mm at 90#{176}C,and then mixed
ical components of muscle described in equation 2. Of with 0.5 ml of 70% PCA. The samples were centrifuged
these six components, we found in preliminary expeni- at 4000 x g for 20 mm and the supernatants collected.
ments that muscle concentrations of NCP, collagen, and The pellets were washed with 1.5% PCA (I x 5 ml) and
TL were stable in situ at 20#{176}Cup to 6 h postmortem. centrifuged again. The combined supernatants were then
However, muscle glycogen, DNA, and RNA declined analyzed for DNA by the Indole procedure of Hubbard
rapidly by about 40 to 50% in the first 60 mm postmor- et al (19), as modified by Wiener et al. (20).
tem, then stabilized for the next 4 to 6 h. Interpretation D W Muscle H,O and DW were determined on a 5-
of these three substrates therefore focuses on the relative ml aliquot of the homogenate. The sample was placed in
difference between groups, rather than absolute levels a small glass vial, frozen at -20#{176}C for 24 h, and then
within each group. Moreover, their total contribution to lyopholized for 48 h. The dry residue was then weighed
wet muscle mass is less than 2%. Muscle energy content immediately.
was calculated from the observed glycogen, and thus will Energy content. Caloric content of muscle (kilocalo-
be low by about 1%. ries) was evaluated by two methods. Method 1 deter-
Muscles were thawed, and the tissue was finely mined kilocalones of dry muscle by bomb calorimetry
minced and homogenized in 25 ml of distilled water. (21) (Parr Adiabatic Bomb Calorimeter, model 1340,
Aliquots were then used for biochemical analysis as Moline, IL). In method 2, muscle energy was calculated
described below. from equation 4, given the measured content of muscle
134 HEYMSFIELD ET AL.

protein, lipid, and glycogen. Results of both methods Relation between muscle mass and
agreed within 2 to 3%, and we, therefore, present data composition
for method 2 only, because this approach allowed
subfractionation of total energy into each of three re- There were no significant differences in the
spective components. composition of psoas and biceps muscles in
the control group (Tables 3A and B), and
Histologic methods
therefore results for all groups are presented
At the time of autopsy, a 10-mg muscle strip was
as “muscle” unless otherwise specified.
removed from the belly of each muscle for histological
study. After fixation in 10% formalin, the slide was Control muscle mass and biochemical com-
prepared by hematoxylin and eosin staining, and exam- position. Arm muscle area and approximate
med by conventional light microscopy. total muscle mass in controls were 48 cm2 and
29.2 kg, respectively. The composition of this
Results muscle tissue, expressed in kilograms and
percentage (Table 3A ) was 23. 1 kg of H2O
Patient profile (79.2%), 4.4 kg of NCP (15%), 1.0 kg of
The 5 1 subjects who met the selection cri- collagen (3.4%),
kg of TL (1.6%),
0.5 0.04 kg
teria are described in Table 2. The control of DNA (0. 12%), 0. 1 kg of RNA (0.35%), and
subjects, seven men and four women, died 0. 1 5 kg of glycogen (0.5%) (Fig. 1).

Downloaded from www.ajcn.org by guest on July 15, 2011

within minutes after traumatic injury. The Total and utilizable muscle energy content
average age and %IBW in this group were 30 was 1.2 and 1.0 kcal/g wet weight, respec-
yr and 103%, respectively. Serum albumin tively (Fig. 2). There were no significant
levels ( = 4.3) were all within our laboratory pathological or histological changes detected
normal range of 3.5 to 5.0 g/dl. by light microscopy.
Patients with early semistarvation, three Muscle in early semistarvation. Compared
men and three women of an average age of to the control group, muscle in early semi-
46 yr, died from cerebrovascular accidents (n starvation showed no detectable difference
= 4), myocardial infarction (n = 1), and auto with respect to mass, composition, or histol-
injury (n = 1). There had been no previous ogy, except for a significant reduction in mus-
history of undernutrition ( %IBW = 113%). dc glycogen and RNA by about 50 to 70%
Chronic illness, usually hypertension, had (p < 0.05). Total and utilizable energy con-
preceded the acute fatal event by an average tent of muscle was 1 .32 and 1.07 kcal/g,
of 8.2 yr, and hospitalization was short, last- respectively, and did not differ significantly
ing only 1 to 4 days ( = 1.6), during which from controls (Fig. 2). Histological examina-
time weight loss averaged 0.4 kg. Serum al- tion revealed no significant abnormalities.
bumin levels ( = 4.3) were within the normal Muscle in chronic semistarvation. Muscle
range. composition in cancer and noncancer patients
The 14 cancer patients with chronic scm- was qualitatively similar, and pooled results
istarvation (12 men and two women) aver- are therefore presented. The mass, chemical
aging 55 yr of age, lost 16.8 kg before and 4.9 composition, energy content, and histological
kg during hospitalization, and were at au- appearance of chronically semistarved mus-
topsy 78% IBW. The average disease duration dc differed from controls. Relative to the
was 3.2 yr. with a final hospitalization of 37 control group, arm muscle area was smaller
days. Serum albumin levels were depressed by 54.2% (this equates to about 14.7 kg of
( = 3. 1 g/dl). The 20 noncancer chronically muscle), but not all components of muscle
semistarved patients (12 men and eight were reduced to the same extent: the relative
women) suffered from cardiopulmonary (n reductions were, respectivley, H2O (1 1.3 kg
= 1 1), liver (n = 8), and renal (n = 1) disease. or 52.9%), collagen (0.4 kg or 46%), NCP (2.7
These subjects tended to be somewhat older kg or 65.3%), TL (0. 1 kg or 40%), DNA (0.0 18
( = 61.7 yr) and less undernourished than kg or 54.1%), RNA (0.021 kg or 81.7%), and
cancer patients. The average total weight loss glycogen (0.005 kg or 90.3%). Since each term
was 12. 1 kg over 7.2 yr, with an additional in equation 3 changed to a different degree,
weight loss of 2.5 kg during hospitalization; the resulting composition of chronically scm-
the average %IBW was 88%. Serum albumin istarved muscle per unit mass differed from
levels were low ( = 3.0 g/dl). controls: H20 (p < 0.05), TL (p < 0.05), and
 MUSCLE COMPOSITION IN SEMISTARVATION 135

I U -
ri.,.
O 00 1- (‘1 *r
o.ri
r-..
r r1 - - -
0. n

.u - tO .F- -

E ‘ t- o ,.
E - (51_

P
-C

:g ;;

r’ r 00 00
- -- t-- 00- -
-

S
u. rr4 r.1

Downloaded from www.ajcn.org by guest on July 15, 2011

r405

- S
V C
U
C c 5. . r,.t
U CO ri-
2

If
2

. #{176}.E’
-©
riJ 1’1 2 a,

9 C rn -. ir r #{231}
. L)
ce-u
n ‘ .oo
0
an ‘#{176}o. -2:’
a, .- #{176}. - N r’ - r’ e’c (‘1
2 #{176} c’cr-4 -- rir .-..

0
.

. E r1 0
‘u L -

.
- ,(

cl
.9. .V 1- (“ r1 r-i 00 E0
c - - en 2’-

c
u
U

_
- an ‘ C ei Q ‘r 00 r-i - Os o
2 < ret - r%l ‘r - C - In - -
0
#{149}:
.a II
Co #{149}0
. C C a
a #{149} .
- 0 ce ce
. n ‘i- II
a - ce C ce
II . #{149}r
‘.- a Ec E11 ?.#{149}9#{149}
r.O . I ‘;: a a
u - . .

0
< a
E- () rJ )
136 HEYMSFIELD ET AL.

* *

* * * * * * *
* * ,r,v_b- * * *

Q
+I-+l 4-I +1 H-+I- 1 41

* *

c & : :
.-

+I’41 l-I +1 414141 41 41 41

* * *
* * * * * * *
* * * * *
C r-
.
+I-+1 .11 I1 41-41-41 .41 44 41
.n n
E
g
. z
* * * * * * -u
. C. * * (..1 ‘r ‘r * * * a,

Downloaded from www.ajcn.org by guest on July 15, 2011

L? .
41_44 f1 .11 41-41-41 41 41 +1 a
0
a

*
* : :
* * - (1 ‘r r’ rt * *

(1 05 ( 0 0 0 0 r - (4 ri - - -
41_41 41 41 414141 41 41 +1
-u

( c6

5.
*
*
:
*
,
* * * *
. . - r’ rt * * *
.- ,

n -

#{176}041_41 -H 41 414141 41 41 41 a

* >
. * *
-r-1r1oo 0
r.i #{176} fit5
. a
41 N44(441 .I1 41 414141 41 +1 413
a,
.n
E
: :
. a, & C. oor-oo---r.i- 0 et ri o o c-i ‘-a,
n 0
. r-44r944 I1 41 44rl44-44 44 44 41_i

-9
ci-

E -r-ir-ooo
-u -
2 r-41t511 41 41 414141 -H 41 IIQ
- 0
-u 2 -u
a>., C
#{176}
a
,-
0 (4-’( r1
c 0
. .
r-44rl44 41 +1 414141 41 41 41 -

ce
a 0
- V
) ci,
. I) *
E -u *
a, C. a . *
Iu- a,
._J0 0 ,,, *

Q
E- ( f-i Z U F- F-
 MUSCLE COMPOSITION IN SEMISTARVATION 137

#{176} ‘,I,, 41 41
* * 41414141-41 * 41
*

ICI *
* *

I&rl * r-1 *
000000e-4
11 I1 41oo11.Hr.441-41

C 0 * * * **
.
0 II 11 11 4111’#{176}11-41-41 I1

. nI
E I
CI
‘) CI
.- ZI *
‘r * c1
. 8.
I.n +1 41 4141#{176}414141 11 ci,
a

Downloaded from www.ajcn.org by guest on July 15, 2011

0
a
*
*

N * * : *

I 9 I1 +3 41oo+ILf1H1r’41-41 11 -

a,
-u
CI
*1 (I
Ut

P 8. b0o”**
r.ir.1
-H
*
*

I1
*
- *

11-a a
.0

. N 41
‘-oo’O0r-1oo 41
* 41
* 4141_41_41
O 0rir1ci *41 . ce

0U
a,
C,, I *

+1 I -“-- 05 r -;
: ‘

0
.- f1 41 11#{176}#{176}111141 41
.
-u
n
a, I r- 2
.
a, 11 +1 11#{176}#{176}11#{176}11-41 +1413
-
-‘I 0
E I
I 5. c - r, N OC -ua
c 11 I1 41oC41N+111 +1 +1i-
0
. I
I

: ci,
,,E -u
a, C. a a
o D
.< .< 3 n . -, *_

;g z L a
F- Q F-’ Z U F- F-
138 HEYMSFIELD ET AL.

CP (p < 0. 1) were increased, NCP (p < 0.01), equivalent to dry control muscle.
RNA (p < 0.01), and glycogen (p < 0.01) The relative change in the composition of
were decreased, and DNA was unchanged. chronically semistarved muscle leads us to
Total and utilizable muscle energy was re- focus on the primary question of the study; is
duced to 1.07 and 0.9 kcal/g wet weight the loss in muscle protein and energy content
(- 10% compared to controls, p < 0.01) (Fig. in proportion to the reduction in muscle mass
2), most of the reduction brought about by in chronic undernutrition? Arm muscle area
increased tissue H20; dry muscle from chron- was 542% smaller than controls (Table 4),
ically semistarved patients was calorically whereas TP, nonstructural protein (NCP) and

100
90
80 H20
4- 70
C.D Non-collagen Protein
w
60
4-
50
w . Collagen Protein
40

Downloaded from www.ajcn.org by guest on July 15, 2011

30 D DNA, RNA Glycogen
20 Total Lipids
10 Extra H20

100
90
80 EElTotal Lipids - Triglyceride
70
C.D
IllI Triglyceride

DNA, RNA, Glycogen

20
w
C-, 15
C,)

-‘ 10
4-
0
4-
5

Control Early Chronic

SEMI-STARVATION
FIG. I . Muscle tissue composition of control group, early and chronically semistarved patients. Results are
expressed as percentage of wet weight (top), percentage of dry weight (middle), and as total muscle mass.
 MUSCLE COMPOSITION IN SEMISTARVATION 139

relations (all p < 0.05) for all three regression

kcal/g WW
analyses.
14 Histological examination of chronically
1.3 semistarved muscle demonstrated variable
1.2 pathology, but generally there was reduction
in myofiber diameter compared to controls,
1.1 and in some cases there was a relative in-
19 crease in connective tissue, interfibrillar
08 edema, and increased fat.
07
0.6 Discussion
05
0.4
Evaluating an array of biochemical muscle
0.3
constituents in a large number of normal and
0.2
undernourished human subjects required
0.1
conditions that must be discussed before in-
00
terpreting our results. The first problem was
Early Chronic the postmortem degradation of RNA, DNA,

Downloaded from www.ajcn.org by guest on July 15, 2011

SEMI-STARVATION and glycogen allowed us to present only rel-
E] Glycogen ative muscle differences of these substrates
Total Lipids - Triglyceride between groups; biopsy and rapid freezing
techniques are required to measure accu-
E1 Triglyceride rately the tissue levels of these compounds in
living man. The second problem was that the
. Collagen Protein
three groups differed in average age. Control
Non-collagen Protein subjects were youngest, while chronically dis-
FIG. 2. Total and utilizable energy per gram of wet eased patients were approximately two dcc-
muscle tissue in control, early, and chronically semi- ades older. Thus the influence of aging itself
starved patients. The proportion of each of the three must be considered in interpreting our results,
energy yielding substrates-protein, lipid, and glycogen as senescence is also associated with muscle
is shown for each group. The broken line depicting
atrophy (22). An important follow-up study
glycogen reflects the variable nature of this substrate,
and the inability in the current study to obtain accurate would, therefore, examine the relation be-
in vivo measurements. The plotted values assume control
glycogen is 1.5% of wet muscle weight, and adjust semi- TABLE 4
starved groups as observed value/control x 1.5%. Percentage change in muscle mass (arm
muscle area) and composition in
chronically semistarved
muscle energy were smaller by 6 1 .6, 65.3, and group relative
59.6%, respectively. Anthropometry alone, to
therefore, underestimated the relative reduc- controls

tion in muscle protein and energy in the Chronic

chronically semistarved group by 5 to 15%. semistarvation

(%
Correlation of serum albumin and muscle Arm muscle area
(equation 5) -54.2
composition. Serum albumin can be related to
muscle composition in two respects. The first Muscle composition
is that plasma oncotic pressure is closely re- (equation 3)
H,O -52.9
lated to the intravascular concentration of CP -46
albumin; low levels lead to leakage of H20 NCP -65.3
into the intercellular space, and the result is TL -40
edema. The second relation is that both NCP DNA -54.1
RNA -81.7
and serum albumin are sources of metabolic
Glycogen -90.3
fuel during semistarvation. We, therefore,
correlated serum albumin with muscle DW Muscle energy
(H20 = muscle mass - DW), NCP and TP (equation 4)
Total kcal -59.6
(Table 5, Fig. 3), and found significant cor-
140 HEYMSFIELD ET AL.

TABLE 5
Correlation of serum albumin (SA) with muscle H,O (DW) and protein*

C Equation SEEt r p

Muscle: biceps
29 SA = 0.016 (DW) + 0.31 0.41 0.5 <0.001
32 SA = 0.019 (TP) + 0.24 0.38 0.7 <0.003
32 SA = 0.02 (NCP) + 0.77 0.36 0.7 <0.0001

Muscle: psoas
32 SA = 0.018 (DW) - 0.29 0.39 0.5 <0.001
32 SA - 0.012 (TP) + 1.37 0.44 0.4 <0.05
28 SA = 0.019 (NCP) + 1.06 0.39 0.56 <0.001
* Units: DW, TP, and NCP mg/g wet wt; SA in g/dl.

t SE of the estimate.

9U unpublished data). Another consideration is

that prolonged attempts at resuscitation
might alter muscle composition in the early
and chronically semistarved patient groups.

Downloaded from www.ajcn.org by guest on July 15, 2011

. Extra H20
Again, other than the loss of liable substrates
D H20 RNA and glycogen, muscle tissue from pa-
tients with early semistarvation closely resem-
D Dry Weight -
bled that from the nonresuscitated control
Total Protein
I Collagen Protein cases. Thus, while all of these factors must be
considered in interpreting our data, it appears
clear that nutritional factors can account for
2.5 3.0 3.5 4.0 4.5 5.0 most ofthe differences in muscle composition
between groups.
SERUM ALBUMIN (601)

FIG. Muscle composition

3. per unit wet weight de- Muscle size and composition in semistarvation
veloped from regression equations in Table 5. If it is
assumed that muscle is normally about 79% H,O, then Early changes. Early semistarvation was
the amount (in percentage) of “excess” tissue water can characterized by a 50% loss in glycogen, but
be calculated as 100 - (4.76 x DW). The excess water no detectable change in muscle size was seen.
line is included in the figure.
We expect, however, that very careful mea-
surements of muscle size in living subjects
tween muscle mass and composition through would indeed detect a reduction in arm mus-
the life span in normally nourished individ- dc area with glycogen depletion for the fol-
uals. Two findings in the current study and lowing reason. For each molecule of glycogen
a parallel rat experiment make it unlikely metabolized, several molecules of associated
that aging muscle alone accounts for all of H20 are also removed from the muscle fiber
the observed differences in muscle composi- (23). The theoretical decrease in muscle size
tion between chronically semistarved and can be calculated if we assume that 1) normal
control groups. The first finding was that in in vivo muscle is 1.5% glycogen/g wet weight
early semistarvation, patients were interme- (254), and 2) muscle remains 79% H20 and
diate in age ( = 47) between the control (5 21% DW irregardless of glycogen content.
= 30) and chronically semistarved groups ( Based on these two assumptions, a 50% loss
= 57); muscle composition, however, was in glycogen would result in a 3.6% reduction
nearly identical to controls, other than the in muscle size, 0.75% of which is glycogen,
expected changes in nutrition-sensitive sub- and 2.85% H20. The 1:3-4 ratio of glycogen
strates RNA and glycogen (discussed below). to H20 is nearly identical to that found cx-
The second observation was that experimen- perimentally in living subjects (25). Similar
tal chronic semistarvation in the rat produced glycogen-dependent changes in arm muscle
biochemical changes in muscle nearly iden- area, but in a positive direction (Fig. 4), would
tical to those observed in the chronically scm- be expected during nutritional recovery when
istarved patients (Stevens V, Heymfield S, there is an increase in muscle glycogen stores.
 MUSCLE COMPOSITION IN SEMISTARVATION 141

AMA 5.26 x DW; “extra H20” for a given DW is

+10 - therefore 5.26/4.76, or 10.5%. The cause of
+5-
excess water accumulation (1 to 1.5 kg in 14.5
%
kg of chronically semistarved muscle, Fig. 1)
-10 -
is unknown, but at least two factors are sug-
gested by our results. The first explanation is
that the decline in serum albumin level
caused a loss of intravascular oncotic pres-
sure, and this allowed H20 to escape into the
0H20
intercellular fluid space. The strong correla-
DRY WEIGHT tion between serum albumin and muscle DW
and H20 suggests this as one possibility. An
additional factor that might have increased
the H20 content of chronically semistarved
muscle was that this tissue contained rela-
tively more collagen, and this protein has a
higher water binding affinity than most of
GLYCOGEN (%WW)
the other cellular proteins (26). Other likely

Downloaded from www.ajcn.org by guest on July 15, 2011

FIG. 4. The manner in which glycogen from control factors tending to expand total body (and
group (CTRL) (assumed to be 1.5% of wet wt) to early
tissue) H20 include endocrine (27), pharma-
(E) and chronic semistarvation (CHR) influences muscle
mass or arm muscle area is shown. Glycogen loss is
cological, and renal mechanisms.
associated with water loss, and the two together account Although muscle H20 accumulation
for about a 3 to 5% theoretical decrease in arm muscle caused the largest discrepancy between mus-
area (AMA). dc size and protein-energy content, there was
also a reduction in TP, and especially NCP/
Chronic semistarvation. Muscle size was re- g of dry muscle weight. Most of the relative
duced in patients suffering chronic semistar- decrease in TP can be accounted for by a 3 to
vation, the reduction brought about by losses 4% increase in TL, especially the TG (Table
of all seven measured components (equation 3A ) lipid faction. Fat replacement of myofi-
3): H20, collagen, NCP, TL, DNA, RNA, brilles (5), or an inability to adequately trans-
and glycogen. Each of these muscle constit- port TG from muscle to other tissues (28),
uents was relatively reduced to a different may account for this finding. Finally, the
degree, and the final composition of the atro- functional protein mass of muscle (NCP) de-
phied muscle differed from controls; relative creased more in chronic semistarvation
losses were l glycogen > RNA > i NCP (65.3%) than muscle size (54.2%), energy
> l\ muscle size = i DNA > collagen> (59.6%), and TP (6 1.6%). The excess H20 and
H20 > i TL. The most important conclu- TG accounted for most of the difference be-
sion from these findings is that muscle size is tween size and z NCP, while the balance
not an exact indicator of muscle protein and was caused by the relative excess of connec-
energy depletion in chronic semistarvation. tive tissue protein.
The average disparity between size and com-
position in our study was not very large- Measurement of muscle mass in
about 5 to 15%, although in some individual semistarvation: clinical perspective
cases the differences were up to 20%. Our study has demonstrated that measure-
The most significant cause of size-compo- ments of muscle mass or size do not neces-
sition dissociation was a relative increase in sarily reflect underlying muscle composition.
muscle H20. If we again assume normal mus- In vivo methods for analyzing the biochemi-
dc 5 79% H20, then excess H20 “falsely” cal composition of muscle are either available
enlarged arm muscle area by 1 to 2 cm2, or (29) or under development (30). Currently,
by about 10%. This calculation is based on however, there is no replacement for the sim-
the following: 1 g of normal muscle = 21% plc and practical bedside anthropometric
DW, or l.0/021 x DW (i.e., 4.76 x DW); if techniques for quantifying muscle mass.
muscle is 19% DW, the equation becomes These methods allow the practitioner to
 142 HEYMSFIELD ET AL.

roughly classify severity of undernutrition ance with different proteins. Anal Biochem
1977;78:257-64.
and establish trends over time. Although we
14. Folch J, Lees M, Stanley OHS. A simple method for
offer no simple correction for the composi- the isolation and purification of total lipids from
tional changes observed in the current study, animal tissues. J Biol Chem l957;226:497-509.
their appreciation should lead to a more ra- 15. Sardesai VM, Manning JA. The determination of
tional interpretation and use of anthropomet- triglycerides in plasma and tissues. Clin Chem
1968; 14: 156-61.
ncdata. U 16. Shibko 5, Koivistoinen P, Tratnyek CA, Newhall
The authors acknowledge the assistance of Eugene AR, Friedman L. A method for sequential quanti-
Semple in collecting the muscle tissue, and Dr. Bob tative separation and determination of protein,
Byers in performing the statistical analyses. RNA, DNA, lipid, and glycogen from a single rate
liver homogenate or from a subcellular fraction.
References Anal Biochem 1967;19:514-28.
17. Carrol NV, Longley RW, Roe JH. The detenmina-
1. Behnke AR, Wilinore JH, eds. Application of the tion of glycogen in liver and muscle by use of
various field methods. In: evaluation and regulation anthrone reagent. J Biol Chem l956;220:583-93.
of body build and composition. Englewood Cliffs, 18. Kamali M, Manhouri H. A modified orcinol reaction
NJ: Prentice-Hall, Inc. 1974:58059. for RNA determination. Chin Chem 1969;5:390-2.
2. Butterworth CD, Blackburn GL. Hospital malnutni- 19. Hubbard RW, Matthew WT, Dubowik DA. Factors
tion and how to assess the nutritional status of a influencing the determination of DNA with indole.
patient. Nutr Today lnc 1975;10:8-18.

Downloaded from www.ajcn.org by guest on July 15, 2011

Anal Biochem l970;38:190-201.
3. Heymsfield SB, Bethel RA, Ansely JD, et al. Enteral 20. Weiner SL, Urivetzky M, Lendvai 5, Shafer 5, Mcii-
hyperalimentation: An alternative to central venous man E. The Indole method for determination of
hyperalimentation. Ann Intern Med 1979;90:63-71. DNA: conditions for maximal sensitivity. Anal Bio-
4. Heymsfield SB, McManus C, Smith J, et al. Im- chem l976;71:583-7.
proved equations for predicting arm muscle area. 2 1. Southgate DA, Dunnin JV. Calorie conversion fac-
Am J Clin Nutr (in press). tons. An experimental re-assessment of the factors
5. Montgomery RD. Muscle morphology in infantile used in the calculation of the energy value of human
protein malnutrition. J Clin Pathol 1962;15:5l1-19. diets. Br J Nutr 1970;24:5 17-35.
6. Heymsfield SB, Olafson RP, Kutner MH, Nixon 22. Safiq SA, Lewis SO, Dimino LC, Schutta HS. Elec-
DW. A radiographic method quantifying protein- tron microscopic study of skeletal muscle in elderly
calorie undernutrition. Am J Chin Nutr 1979; subjects. In: Daldor G, DiBattista Wi, eds. Aging.
32:693-702. 3rd ed. New York: Raven Press, 1978:65-85.
7. Blackburn GL, Bistrian BR, Maini BS, et al. Nutni- 23. Bergstrom J, Hultman E. Synthesis of muscle gly-
tional and metabolic assessment of the hospitalized cogen in man after glucose and fructose infusion.
patient. JPEN 1977:1:11-22. Acta Med Scand l967;l82:93-107.
8. National Research Council: Food and Nutrition 24. Bergstrom J. Muscle electrolytes in man. Scand J
Board. Recommended dietary allowances. 9th ed. Clin Lab Invest 1962;14(suppl 68):42.
Washington, DC: National Academy of Sciences, 25. Grande F, Keys A. Body weight, body composition
1980. and calorie status. In: Goodhart RS, Shils ME, eds.
9. Begg RW, Dickinson TE. Systemic effects of tumors Modern nutrition in health and disease. 6th ed.
in force-fed rats. Cancer Res 195 1;1 1:409-12. Philadelphia: Lea and Febiger, 1980:3-35.
10. Boyd MB, Connell ML, McEwen HD. The lipid 26. Kuntz ID Jr. Kauzmann W. Hydration of proteins
composition and water content of carcass, skeletal and polypeptides. Advan Protein Chem 1974;
muscle, and testicle in the host component of the 28:239-345.
albino rat-Walker carcinoma 256 dual organism. 27. Hansen JDL. Endocrines and malnutrition. In: 01-
Can J Med Sci 1952;30:47 1-83. son RE, ed. Protein-calorie malnutrition. New York:
I 1 . Boyd MB, Boyd CD, Hill JO, Ravinsky E. The lipid Academic Press, 1975:229-41.
and water content of carcass, skeletal muscle, and 28. McGarry JD, Foster DW. Systemic carnitine defi-
testicle in the host component of the albino rat- ciency. N Engli Med 1980:11:1413-15.
Walter carcinoma 256 dual organism at progressive 29. Collins JP, McCarthy ID, Hill GL. Assessment of
stages of tumor growth. Can J Biochem Physiol protein nutrition in surgical patients-the value of
1954;32:359-7 1. anthropometrics. Am J Clin Nutr 1979:32:1527-30.
12. Lilienthal JL Jr. Zierler KL, Folk BP, Buka R, Riley 30. Mallard J, Hutchinson JMS, Edeistein WA, Ling
Mi. A reference base and system for analysis of CR, Foster MA, Johnson 0. In vivo N.M.R. imagery
muscle constituents. J Biol Chem 1950;l82:501-16. in medicine. The Aberdeen approach, both NMR
13. Dorsey TE, McDonald PW, Roels OA. A heated physical and biological. Philos Trans R Soc Lond
biuret-folin protein assay which gives equal absorb- [Biol] l980;289:5 19-33.

View publication stats