Method for estimating body weight in persons with lower-limb amputation and its implication for their nutritional assessment1-3
A r u p e n d r a M o z u m d a r a n d S u b r a t a K R o y
A BSTRA CT
Background: Body weight is a good indicator of a person’s size and is widely used in clinical assessment. However, health-status assess
m ent based on observed body weight (W0 ) is incorrect for persons with limb amputation.
Objectives: The objectives were 1) to develop a m ore accurate and generalized method for estim ating body weight in persons with limb amputation, 2) to determine whether corrected body weight can be used to assess nutritional status in persons with limb amputation, and 3) to test the validity o f the estimation by using empirical data.
Design: Anthropometric data were collected from men from Cal
cutta and adjoining areas w ith unilateral lower-extremity amputation (n = 102). Mathematic formulas were developed for determining estimated body weight (W E) and body mass index (BMI) calculated from both WQ and WE (ie, B M I0 and BMIE, respectively). We assessed nutritional status by using BMI0 and B M IE and tested the validity of each by considering the result of nutritional assessment from midupper arm circumference as the gold standard. We also com pared the nutritional status results for the subjects with limb amputation with those for a similar sample size o f healthy control subjects.
Results: BMIE had a stronger association with m idupper arm cir
cumference and a higher efficiency (ie, proportion o f correct results given by any test method) than did BMI0 . Moreover, the results obtained with BMIE w ere similar to those obtained with BMI in healthy control subjects. H owever, the nutritional assessments made with BM I0 and BMIE did not differ significantly from one another.
Conclusion: For persons w ith limb amputation, WE provides a better basis for appropriate nutritional evaluation than does WQ. A m J Clin Nutr 2004;80:868-75.
K EY W ORDS B ody weight, body proportions, persons with lower-limb amputation, anthropometry, body mass index, nutri
tional assessment
INTRODUCTION
B o d y w eight generally reflects m any physical attributes (size a n d shape) o f the h u m a n body. D espite significant variation attributable to sex, stature, age, and socioeconom ic conditions, b o d y w eig h t is often u se d as an indicator of the n utritional status an d m o rb id condition o f a p erso n because o f its sensitivity to e nvironm ental conditions. T herefore, body w eig h t is an im por
ta n t physical characteristic th a t can be helpful in m a k in g clinical assessm ents, including d eterm in in g the appropriate dosage o f m e d icin e and appropriate fo o d supplements.
F o r h ealth y persons, body w eig h t can be estim ated very easily and accurately. H owever, bo d y -w eig h t estim ation becomes crit
ical a n d som etim es com plicated in the case o f persons who are dev o id o f a lim b or a part o f a lim b u su ally due to amputation or c ongenital defects. For persons w ith lim b amputation, body w eig h t obtained by using the stan d ard m ethod is generally an u n d erestim ate and does not properly reflect their body shape and size. T h e re are 2 alternatives for estim atin g the body weight of a p erso n w ith an am putated limb: 1) w eigh the am putated portion o f th e b o d y (at the tim e of the am p u tatio n ) and then add the extra w eight, w h ic h is hardly practical; or 2) estim ate the weight of the am p u tated portion o f the body fro m the observed body weight (po stam p u tatio n ) by using b o d y -w eig h t proportions.
P a st studies on body-w eight p ro p o rtio n s were based on mea
surem ents o f the w eight o f separated b o d y segm ents from human cad av ers (1 -3 ). In 1964 H anavan (4) d eveloped a computerized seg m en t m o d e l o f the hum an bo d y w ith the use o f 25 anthropo
m etric m easurem ents. Several lim itatio n s o f the earlier studies have b e e n pointed out. These lim itatio n s include the fact that g en eralizatio n o f the results m ay b e d ifficult because o f variation in b o d y -w e ig h t proportions due to sex (5, 6), age (7-10), and ethnicity (1 1 ,1 2 ). H owever, "Wilson and L oesch (13) showed that the sh ap e variables o f trunks and lim b s in both sexes are similar.
M a r to re ll e t al (14) fo u n d th a t a l th o u g h le n g th m easurem ents are a ff e c te d b y so cio eco n o m ic s ta tu s su c h th a t p o o r persons are lik e ly to b e short, so c io e c o n o m ic fa c to rs do not affect r e la tiv e b o d y p ro p o rtio n s, an d th is fin d in g has b een confirmed in o th e r s tu d ie s (15). T an n er (1 5 ) a lso o p in e d th a t the body p ro p o r tio n s o f E u ro p e a n and A s ia tic p o p u la tio n s are similar b u t d if f e r fr o m th o se o f A fric a n p o p u la tio n s . H ow ever, all th e se s tu d ie s w e re b ased on a s s e s s m e n t o f the relativ e size of the b o d y se g m e n ts , not o f th e a c tu a l w eig h ts o f the body s e g m e n ts .
T h erefo re, the need for a sim ple m e th o d for estim ating body- w eig h t p roportions in persons w ith lim b am putation was felt, alth o u g h the ex act assessm ent o f th e id e al proportion o f the body w eig h t w h e n all possible varying factors are considered is very
1 From the A n th rop ology and H um an G en etics Unit, Indian Statistical Institute, Calcutta.
2 Supported by the Indian Statistical Institute, Calcutta.
3 Reprints not available. Address corresp on d en ce to SK Roy. Anthropol
o g y and Hum an Genetics Unit, Indian Statistical Institute, 203, Barrackpore Trunk R oad, K olk ata 7 0 0 108, India. E-m ail: [email protected].
868 A m J C lin N u t r 2 0 0 4 ; 8 0 : 8 6 8 - 7 5 .
complex. An attempt was recently made to generalize the find
ings at'published studies based on the actual assessment of seg
mental body weight, coupled with trends in generational change, which also supported the utility of that generalization of the results for the assessment of nutritional status (16). In the course ol review ing the w ork of Osterkamp (16). an additional applica
tion of nutritional assessment by formulating the body mass index (BM1: in k g /n r) of persons with limb amputation was pointed out. An effort was also made to estimate total body weight (ie. before amputation) by using a mathematic model for persons with amputation through the knee only (17). However, in reality the frequency of amputation through the knee is much lower than that of amputation above or below the knee, in which the limb loss is transfemoral ortranstihial.Tzamaloukaset al ( 18.
19i also developed a mathematic model for predicting the esti
mated (preamputation) weight of persons with limb amputation and attempted to estimate the body mass index and nutritional status of persons w ith limb amputation. However, the studies by T/amaloukaset al had some limitations, such as the following: 1) the) used an outdated method (20) for estimating the preampu
tation weight of the persons with limb amputation, and 2) their sample sizes were very small and heterogeneous.
B.M1 is often used as an indicator of nutritional status inhuman populations (211. However. BMI has rarely been used for the assessment of nutritional status in persons with limb amputation.
BMI was used in a study in Israel as one of the obesity indexes to assess cardiovascular health in persons with limb amputation due to trauma 122). Some studies of physically disabled persons have also used BMI. although data from persons with limb amputation have nut been analyzed separately: for example, few studies have included both aged and disabled persons (23. 24), and studies have also calculated BMI in disabled children (25). The reason behind the limited number of studies may be the problem in measuring bodv weight and stature in persons with limb ampu
tation. Although stature measurement is possible to some extent in the ease of persons with unilateral amputation, the correct body weight (proportional to body shape and size) is difficult to mea
sure because of the loss of proportion of body weight due to limb amputation. This ultimately leads to an underestimation of BMI as well as an underestimation in nutritional assessment. In view of the need for efficient estimation of body weight in persons with limb amputation and the limitations of earlier studies as stated above, the purposes of the present study were /) to develop an accurate method for estimating body weight in persons with limb amputation. 2) to determine whether corrected body weight can be used to assess nutritional status in persons with limb ampu
tation. and .?) to test the validity of the estimation procedure by using empirical data from persons with lower-extremity ampu
tation from Calcutta and its adjoining areas.
SUBJECTS AND METHODS Population and a r e a
The data used in the present study were collected as part of a larger biomedical program involving persons with lower- extremity amputation from Calcutta and its adjoining areas. Two national-level rehabilitation centers, the National Institute for the Orthopedically Handicapped and Mahavir Seva Sadan, were :ontacted for a list of addresses of persons with lower-extremity imputation. A statement of purpose of the present research and
a consent form were mailed to these persons. Respondents, who provided written informed consent, were included in the study.
The study was performed according to the responsible committee on human experimentation (Scientific Ethical Committee for Protection of Research Risks to Humans, Indian Statistical In
stitute). Data were collected from a total of 102 men with uni
lateral lower extremity amputation; 32 of these men had ampu
tations above the knee, and 7 0 had amputations below the knee.
T h e mean ( ± S D ) age of the subjects was 43.54 ± 15.37 y. A large proportion (82.6% ) of the subjects had amputation due to trauma, only 11.0% had amputation due to degenerative disease, and the remaining 6.4% had a reported history of cancer. All the subjects had prostheses, and all of them had been amputated > 2 y before this study. All data were collected by a single investi
gator (AM) through multiple home visits. In addition, data were co lle c te d from a control g ro u p m a de up of 105 healthy m en who w ere m atch ed to the subjects w ith limb am putation b y age and so cio eco m o n ic status. All the subjects with limb amputation and all the healthy control subjects were Bengali-speaking H indus.
A nthropom etric m easurem ents
All anthropometric measurements were performed with the use of standard techniques (26). The subjects with limb ampu
tation were requested to wear a prosthesis before having their stature and body weight measured (if required, the subjects were supported against a wall with adequate precautions to guard against bending of the trunk and knees). The weight of the pros
thesis was then taken and subtracted from the previous weight with the prosthesis to get the actual weight (postamputation) of the body. To our knowledge, there is no standard method for measuring the stature of persons with limb amputation. There
fore, the stature measurements of the person with limb amputa
tion were cross-checked for consistency by calculating body proportions (sitting height/stature) (27) and comparing them with those o f persons without limb amputation.
Analysis
Theoretical considerations
If the present body weight of a person with limb amputation (postamputation) is W0, estimated (preamputation) total body weight is Wv, and the reduction of the body weight due to the amputation is AW, then the following formula can be developed.
W0 = WE - AW (7)
or, dividing both sides by WE
Wo/WE = 1 - AW/WE (2)
or
WE = W0/( l - AW/WE) (5) Now AW/WE is the proportion of the body weight lost due to amputation from the total body weight.
Amputation may be of different components of the limbs (in
cluding partial amputation of limbs), and the weight loss due to these components (a, b , c , . . . , n ) can be denoted as AWa, AWb,
A Wc, A Wn. Therefore, the total weight loss of the body (ie, A W) is
AW = A Wa + A Wb + AWC + . . . + A W„ (4) or
i W E = AWa/WE + A W J We + AWC/WE . . . + AW„/WE (5) The value of AW (ie, the weight loss due to amputation) is not uniform among most persons with limb amputation. For exam
ple, an amputation below the knee results in a partial loss of the leg and a complete loss of the foot. In the case of an amputation above the knee, the weight loss is due not only to the complete loss of the lower leg and the foot but also to a partial loss of the thigh. Similarly, an amputation through the knee results in a weight loss of the lower leg and the foot but no partial loss of any other segment of the lower extremity. Lower-extremity amputa
tions and loss of weight, therefore, may broadly be classified into the following types:
Ankle disarticulation amputation, A W — AWS (6) Amputation below the knee (transtibial), AW = AWS
+ ApWt (7) Amputation through the knee, AW = AWS + AWT
(8)
Amputation above the knee (transfemoral), AW = AWS + AWt + ApWF (9) Hip disarticulation am putation, AW = AWS + AWT + AWF
(10) where AW = total weight loss, AWS = weight loss due to the loss of afoot, A WT = weight loss due to the loss of a leg (tibial region), Ap Wx = partial weight loss due to the partial loss of a leg, A WF = weight loss due to the loss of a thigh (femoral region), and ApWF = partial weight loss due to the partial loss o f a thigh.
According to Osterkamp (16), the proportions of AWS, AWT, and AWF to total body weight (ie, WE) are 1.5%, 4.4%, and 10.1%, respectively. Therefore, the proportion of weight loss (lower-limb amputation) for ankle and hip disarticulation ampu
tations and amputations through the knee are easier to estimate with the use of these proportions from Osterkamp (16). However, in real life, amputations below or above the knee (ie, transtibial or transfemoral), which involve the partial loss of the tibial or femoral region, respectively, are more common. However, no method to estimate the proportion of weight (lost or retained) of the tibial or femoral region is available. It is not feasible to weigh the amputated portion (unless the weighing is done at the time of the amputation), and it is also difficult to know the weight of the stump (ie, the remaining portion of the limb from its nearest distal bone joint). Thus, one has to rely on the proportional estimation of the size of the stump relative to that of the total region (ie, the tibial region in the case of amputations below the knee and the femoral region in the case o f amputations above the knee). How
ever, such estimation procedures require additional anthropo
metric measurements of the persons with limb amputation. The
measurements are I) length of the stump, and 2) knee height foi persons with amputation below the knee and buttock-knee length for persons with amputation above the knee (see reference 26 for technical details). These measurements have been used widely in many ergonomics studies (28, 29) of subjects with amputated limbs. Although, measurement of knee height or buttock-knee length is obviously not possible for the amputated limb, in the case o f persons with unilateral limb amputation, these measures may be replaced by those of the available limb, with the assump
tion of bilateral symmetry. Thereafter, estimation of the remain
ing proportion of the limb may be made by calculating the pro
portion of the length of the stump to the knee height or buttock- knee length as follows:
ApWT = AWt - AWt X L ^ / L Kn = AW,(1 - LSlp/LK„)
(11)
ApWF = AWf - A WF X LStp/LBlK = AWf( 1 - LStp/LBtK)
(1 2)
where LSlp = length of the stump, L Kn = knee height, and Z.BtK = buttock-knee length.
Therefore, new corrected formulas for estimating the lost pro
portions of total body weight in the case of persons with ampu
tations below or above the knee may be written as follows:
A m putation below the knee (transtibial), AW = AWS
+ A W t ( 1 - L StpI L Kn) (13)
or
AW/WE = AWs/We + AWt /We X (l - LSlp/LKn)
(14)
A m putation above the knee (transfem oral), A W = AWS + AWt + AWF(1 - LStp/LBlK) (75) or
AW/We = AWs/We + A WT/WE + AWf/We X (1 - LSip/Lb.k) (16) The proportions AWS/WE AWT/W E, and AWF/WE can easily be obtained from the findings of Osterkamp (16) as mentioned be
fore. Moreover, from the subsequent calculation of A W/WE it is
possible to estimate WE by using Equation 3.
This method of estimating total body weight obviously per
tains to persons with unilateral limb amputation only. For per
sons with bilateral amputation, there is no scope for using tfe measurement of the normal limb to predict the proportion of
weight loss from the amputated limb. Stature measurement is generally used to predict the length proportion of limb segments However, it is impossible to obtain stature measurements of
persons with bilateral limb amputation because of the absence of both lower limbs. Several efforts have been made to develop some methods for measuring the stature of persons with bilateral locomotor disability (28), but these could not be adequately stan- dardized. Therefore, we did not consider stature m e a s u r e m e n t d
persons with bilateral limb amputation. Many studies of persons with limb amputation show that a linear body m e a s u r e m e n t thi
is strongly correlated with stature, such as sitting height (27) 0
Linn s p a n 13* * i. v a n b e c o n s i d e r e d i n s t e a d o f s t a t u r e . I n t h e p r e s e n t s n u b . s i t t i n g h e i g h t m e a s u r e m e n t w a s c o n s i d e r e d f o r c a l c u l a t i n g b o J v p r o p o r t i o n s ; s u c h m e a s u r e m e n t m a y h e o b t a i n e d e v e n f r o m p e r s o n ' w i t h b i l a t e r a l l o w e r - l i m h a m p u t a t i o n w i t h t h e u s e o f sUiih1.ii.-1 t e c h n i q u e s a n d i n s t r u m e n t s . D r i l l i s a n d C o n t i n i ( 3 1 ) d e v e l o p e d a s c h e m a t i c d i a g r a m o f h u m a n b o d y p r o p o r t i o n s b y usin^ l i n e a r m e a s u r e m e n t s . T h e s i t t i n g h e i g h t p r o p o r t i o n o b tained bv vii\ u l m g t h e s i t t i n g h e i g h t b y s t a t u r e i s e s t i m a t e d t o b e O . v i ; | i. a l t h o u g h t h e r e i s s l i g h t v a r i a t i o n a c r o s s e t h n i c g r o u p s .
F r o m t h e a b o \ e . it / . s , i s s t a t u r e a m i / . s , n i s t h e s i t t i n g h e i g h t o f a p erso n , t h e n
/ s / .s V / ’mHV (17)
where / ’v-m s |v p r o p o r t i o n o f s i t t i n g h e i g h t t o s t a t u r e ( i e . 0 .5 21. ' \ \ c l e c o m m e i t i l t h a t t h i s p r o p o r t i o n b e e s t i m a t e d i n d e - penJoiith l o t t i l e s p e c i f i c e t h n i c g r o u p b e i n g s t u d i e d b y t a k i n g m c .i'U ie m e n t s m a g r o u p n f h e a l t h ) p e r s o n s w i t h o u t l i m b a m putation !i> 'in t h a t p o p u l a t i o n . I
E - u n u l t n i i o l t o t a l w e i g h t l o s s in p e r s o n s w i t h b i l a t e r a l l i m b ampiit.iiion c a n t h u s b e d o n e i n t h e f o l l o w i n g w a y . S u p p o s e a person h a s u n d e r g o n e a m p u t a t i o n b e l o w t h e k n e e i n t h e l e f t l e g a n ih iN '\e ' h e k n e e m t h e r i g h t l e g . In s u c h a c a s e , t o t a l w e i g h t loss van h e i c p i e s e n t e d b y t h e f o l l o w i n g e q u a t i o n :
A U A H ' , • A U K (1 8)
where A H , is t h e w e i g h t l o s s f r o m t h e l o s s o f t h e l e f t l e g , a n d AlVk i' the u c i ' j l i t l o s s f r o m t h e l o s s o f t h e r i g h t l e g .
From h q i i a i i o n / • / . t h e f o l l o w i n g f o r m u l a c a n b e d e r i v e d : ill', = A H '. ■ A l l a I / - s :r/ / - K „ l A U . + A U ' . j 1 -
/ . Mp/( P Kl,,s, X L Sl) j (1 9)
where / \ s , is t h e p r o p o r t i o n o f k n e e h e i g h t t o s t a t u r e , a n d L St i s th ep r ea m p u ta lio ii s t a t u r e , f r o m l i q u a t i o n 16 . t h e f o l l o w i n g f o r mulas can b e i l e m e d :
A U ;.. A l l s • A l l i A H | ( I — Z . s ,r / Z . M, K ) (20) All:, A l l s • A l l i • A l l 11 I — L Stp/ ( / >i a \ si L $ , ) ]
(2 1)
where/Jm N S[ i s t h e p r o p o r t i o n o l b u t t o c k - k n e e l e n g t h t o s t a t u r e . Asa n ote o l c a u t i o n . o n K i n c a s e s i n w h i c h it i s i m p o s s i b l e t o measure the b u t t o c k - k n e e l e n g t h o r k n e e h e i g h t f r o m o t h e r l i m b measurem ents , s u c h a s in p e r s o n s w i t h b i l a t e r a l l i m b a m p u t a t i o n , should t h o s e m e a s u r e m e n t s h e e s t i m a t e d I r o m s t a t u r e .
Statistical iintily\i\
D e s c r i p t i v e s t a t i s t i c s w e r e c a l c u l a t e d f o r a l l a n t h r o p o m e t r i c m easu r em en ts f r o m t h e s u b j e c t s w i t h l i m b a m p u t a t i o n ( b e l o w the k n e e , a b o v e t h e k n e e , a n d p o o l e d ) a n d f r o m t h e h e a l t h y control s u b j e c t s w i t h o u t l i m b a m p u t a t i o n . T h e t o t a l b o d y w e i g h t s (p rea m p u ta tio n ) o l t h e s u b j e c t s w i t h l i m b a m p u t a t i o n w e r e e s t i mated a c c o r d i n g t o E q u a t i o n s 1 4 a n d 16. a n d d e s c r i p t i v e s t a t i s tics w e r e c a l c u l a t e d .
The ratio of sitting height to stature w as c a lc u la te d f o r all the subjects with limb am putation a n d fo r the co ntrol s u b je c ts , and descriptive statistics were calculated. F u rth e rm o re , o n e -w a y analysis of variance was p e rfo rm e d to te st w h e th e r th e d if fe re n c e in mean values for the ratio o f sitting heig h t to sta tu re b e t w e e n the
subjects with limb amputation and the control subjects were statistically significant.
BMI was calculated by using both WQ (BMI0) and WE (BMIE).
The terms BMI0 and BMlEhave been used consistently through
out the article.
Correlation coefficients between midupper arm circumfer
ence (MU AC) and the 2 BMI values (ie, both BMI0 and BMIE) were calculated for the subjects with limb amputation. It is worth mentioning that MUAC generally has a strong positive correla
tion with BMI in healthy adult populations and can therefore be used as a measure of nutritional status in adult populations (32).
The nutritional status of the subjects with limb amputation was classified by using cutoffs (33). An MUAC 5: 24.3 cm was considered normal, and an MUAC < 24.3 cm was considered indicative of the presence of chronic energy deficiency (CED).
Similarly, a BMI > 18.5 was considered normal, and a BMI <
18.5 was considered indicative of the presence of CED.
The test of sensitivity and specificity is a way of examining the effectiveness of the prediction protocols. A good predictor is one that has high sensitivity and high specificity, but one must some
times choose a balance between the 2 because it is very difficult to determine which one is more important than the other. Some
times, a test with high sensitivity is desired. For example, when a blood bank tests blood for HIV, a test with 100% sensitivity is desired, although some false-positive test results will occur. Con
versely, before doing an autopsy, a pathologist tests for the pres
ence of death; this test requires high specificity to avoid autop- sying someone who is not dead. However, measuring the effectiveness of a test method that has been developed is always necessary. To test effectiveness, a statistical tool called “effi
ciency” has been calculated; it is the proportion of correct test results given by any test method (ie, true positives + true nega
tives/total number of cases) (34).
It is not always possible to identify a true condition, which is biologically independent of the test method under investigation.
Therefore, a reference test (gold standard) was used to examine the performance of a new test. In such a circumstance, the result of the reference test is considered to be indicative of the true condition, and the efficiency of the new test method is tested accordingly.
MUAC is a good predictor of nutritional status in adults and for screening populations (32, 33). In many instances, MUAC has been used as the first screening method in a crisis setting (35,36) and has been shown to be a good predictor of morbidity and mortality as well (37). MUAC also has a good correlation with BMI (32,33). Therefore, MUAC was used in the present study as the gold standard against which the estimated method was cali
brated.
Agreement statistics (Cohen’s k) between MUAC and BMI0 and between MUAC and BMIE were calculated in the subjects with limb amputation (below the knee, above the knee, and pooled), and agreement statistics between MUAC and BMI were calculated in the control subjects. The « value indicates the de
gree of association, which varies from —1 to 1, and a k value close to 1 indicates a very strong association.
In the present analysis, the sensitivity, specificity, positive predictive value, negative predictive value, and efficiency of both BMI values were calculated for persons with limb amputa
tion and control subjects by using MUAC as the gold standard to determine the validity of the calculated BMI values for the
TABLE 1
D e s c r i p t i v e s tatistics f o r a n th r o p o m e tr ic v a r ia b le s in t h e s u b j e c t s w it h li m b a m p u t a t i o n ( b e l o w th e k n e e , a b o v e t h e k n e e , a n d p o o l e d ) and in t h e control s u b j e c t s '
C o n t r o l subjects Ui = 105) 163.79 ± 6.52
.59.71 ±11.28 8.5.15 ± 3.75 2 7 .1 3 ± 3.29 5 0 .S5 ± 2.77 .54.62 ± 3.1)0
! A ll v a lu es are x ± S D . M U A C , m id u p p e r a rm c i r c u m f e r e n c e .
2 J S ig n ific a n tly d if f e r e n t f r o m co n tro l s u b jec ts ( t test): 2P < 0 . 0 5 , JP < 0.01.
S u b j e c t s w it h l i m b a m p u t a t io n
A n t h r o p o m e t r ic variable
B e l o w k n e e (n = 7 0 )
A b o v e k n e e (n = 3 2 )
P o o l e d
(/i = 102)
S ta ture ( c m ) W e i g h t (k g ) S it t in g h eig h t (cm ) M U A C (c m ) K n e e h e ig h t (cm ) B u t t o c k - k n e e le n g th ( c m )
1 6 0 .8 5 ± 7 . 1 5 5 4 . 4 9 ± 1 1 . 3 6 8 3 . 2 2 ± 3 , 8 3 2 6 . 4 0 ± 3 . 3 7 5 0 . 1 4 ± 3 . 0 7 5 4 . 5 7 ± 3 . 5 3
1 6 4 . 4 2 ± 6 . 8 3 5 9 . 9 7 + 1 2 .4 1 8 4 . 6 6 + 3 . 7 4 2 9 . 1 4 ± 3 . 2 4 5 1 . 9 7 ± 2 . 9 2 5 5 . 7 5 + 3 . 4 9
1 6 1 . 9 7 + 7 .2 1 5 6 . 2 1 + 1 1 . 9 2 - 8 3 . 6 7 + 3 . 8 4 ‘ 2 7 . 2 6 ± 3 . 5 5 5 0 . 7 2 ± 3 . 1 3 5 4 . 9 4 ± 3 . 5 4
present purpose. All statistical analyses were performed by using SPSS for WINDOWS (version 7.5; SPSS, Chicago).
RESULTS
Descriptive statistics for anthropometric traits in the subjects with limb amputation (below the knee, above the knee, and pooled) and in the control subjects are shown in Table 1. When pooled data from the subjects with limb amputation were com
pared with the data from the control subjects, the subjects with limb amputation were found to have significantly lower weight (P < 0.05) and sitting height (P < 0.01). The mean values for all the anthropometric variables were higher in the subjects with an amputation above the knee than in those with an amputation below the knee.
Descriptive statistics for WD and WE, the ratio of sitting height to stature, and body mass index in the subjects with limb ampu
tation and in the control subjects are shown in Table 2. The estimated weight of the subjects with limb amputation was cal
culated by using Equations 14 and 16 (described earlier). The difference between WQ and WE in the subjects with an amputa
tion above the knee was larger than that in the subjects with an amputation below the knee, because this difference includes the weight loss due to amputation, which is greater for the subjects with an amputation above the knee. An attempt to determine the possible effect of age on the calculated ratio (proportion) of sitting height to stature was also made; however, the effect of age
on the variable was found to be negligible. In one-way analysis o f variance, the ratio of sitting height to stature did not differ significantly between the subjects with limb amputation and the control subjects. The mean value of BMI,. was higher than that of BMI0 in each group of subjects with limb amputation because \ WE is always higher than H'<
MUAC was significantly positively correlated with both BM10 (r = 0.846, P < 0.01) and BMI, (r = 0.872. P < O.Oli Moreover, in the present control population, a very strong pos
itive correlation was found between MUAC and BMI as well.
The frequency distribution of nutritional status in the subjects with limb amputation (below the knee, above the knee, and pooled) and in the control subjects is shown in Table 3. Classi
fications were made by using MUAC and BMI cutoffs forCED.
Individual values ofboth estimates of BMI (ie. BMI()andBMIE!
and o f MU AC were considered for cross-tabulation. and frequen
cies of subjects were calculated accordingly. For MUAC pooled (total) data, 26 and 76 subjects had CED and normal nutritional status, respectively, according to BM1(). and 16 and 86 subjects had CED and normal nutritional status, respectively, according to BMIe . Thus, more subjects had normal BMI if BMIE was considered instead of BMI0 (because of the compensation for weight loss in the BMIE calculation); however, some of the.
subjects classified as having normal nutritional status are instead classified as having CED if BM10 is considered. There were no significant differences in BMI distribution between the control
T A B L E 2
D e s c r i p t i v e statis tics f o r o b s e r v e d a n d e s tim a te d b o d y w e i g h t (WQ an d WE, r e s p e c t i v e l y ) , th e ratio o f s ittin g h e i g h t to s t a t u r e , a nd B M I c a l c u l a t e d from H0 a nd tVE ( B M I 0 and B M I E, r e s p e c t i v e ly ) in th e s u b jec ts w it h l i m b a m p uta tio n ( b e l o w t h e k n e e , a b o v e th e k n e e , a n d p o o l e d ) a n d in th e c o n t r o l s u b j e c t s '
S u b j e c t s w ith l i m b a m p u t a t i o n B e l o w k n e e
(n = 7 0 )
A b o v e k n e e (n = 3 2 )
P o o l e d (n = 1 0 2 ) Wc ( k g ) 2
WK (k g )
S ittin g height/stature'1 B M 1 „ (k g / n r f BMI,: ( k i i / n r )
5 4 . 4 9 ± 1 1 . 3 6 5 6 . 6 2 ± 1 1 . 8 2 0 . 5 1 7 5 ± 0 . 0 1 4 2 0 . 9 6 ± 3 . 5 4 2 1 . 7 8 ± 3 .6 7
5 9 . 9 7 + 1 2 . 4 1 6 6 . 3 0 + 1 3 . 6 6 0 . 5 1 4 9 + 0 . 0 1 1 2 2 . 0 0 + 3 . 8 2 2 4 . 3 6 ± 3 . 4 3
5 6 . 2 1 ± 1 1 . 9 2 5 9 . 6 5 + 1 3 . 1 5 0 . 5 1 6 7 + 0 . 0 1 3 5
2 1 . 2 9 ± 3 . 8 9 2 2 . 5 8 ± 3 . 5 2 ' A ll v a lu es are v ± S D .
■ O b s e r v e d b o d y w e i g h t o f th e s u b jec ts w ith li m b a m p u t a t i o n and actual b o d y w e i g h t o f t h e c o n t r o l s u b j e c t s . ' N o s ig n ifica nt d i f f e r e n c e b e t w e e n g ro u p s ( A N O V A ) .
’ B M I „ = BMI in th e c o n t r o l subjects.
C o n t r o l subjects (i! = 105) 5 9 . 7 1 + 11.28
0 . 5 1 9 9 + 0.013 2 2 . 2 0 ± 3.60
TABI.K 1
Coniii)i!cih \ t a b le for d is t r ib u t io n o f the s u b j e c t s w i t h li m b am p uta tio n ( b e l o w t h e k n e e , a b o v e th e k n e e , a n d p o o l e d ) and o f the co n t r o l s u b j e c t s a c c o rd in g to thou n utritio nal s l a w s a s a s s e s s e d b y u s i n g m i d u p p e r arm c ir c u m f e r e n c e ( M U A C ) a n d B M I (in k g / m 2)'
MUAC
B e l o w knee A b o v e k n e e P o o l e d C o n t r o l s u b jec ts
C 1 :. D J N o rm a l* T o ta l C E D N o r m a l T o t a l C E D N o r m a l T o ta l C E D N o r m a l Total
ii n n n
BMI,,
C l - i r 14 1 21 1 4 5 15 11 26 15 5 2 0
N orm al' 4 4 5 4 9 — 27 2 7 4 7 2 7 6 4 81 8 5
Total IS 5 2 7 0 1 31 3 2 19 8 3 1 02 19 8 6 105
BMI,
I I I ) 12 3 15 1 — 1 13 3 16
Nonn.il (> 4 9 5 5 — 31 31 6 8 0 8 6
Total IS 5 2 7 0 I 31 3 2 19 8 3 1 02
C I D . c h r o n ic e [ k 'n j \ d e f ic ie n c y : B M I , , . B M I c a l c u la t e d from o b s e r v e d b o d y w e i g h t ( B M I 0 = B M I i n t h e co n t r o l su b jec ts); B M I E, B M I c a l c u la t e d from estimated b u d \ w e i g h t in th e s u b jec ts w ith lim b a m p u t a t i o n . A g r e e m e n t s t a t is t ic s b e t w e e n M U A C a n d B M I 0 w e r e a s f o ll o w s : k = 0 . 6 1 0 ( i n s u b j e c t s w ith an a mp u t a t i o n b elo w th e k n e e i. (1.297 (in s u b je c ts w i t h an a m p u ta tio n a b o v e t h e k n e e ) , a n d 0 . 5 7 5 (in p o o l e d s u b j e c t s w it h li m b am p u ta tio n ). A g r e e m e n t s tatis tics h e i u e c n M l A C .u ul H M I, w e r e as f o llo w s : k = 0 . 6 4 4 (in sub je c ts w ith an a m p u t a t i o n b e l o w the k n e e ) , 1 . 0 0 0 ( i n s u b j e c t s w it h an a m p u t a t i o n a b o v e the k n e e ), and i (in p o o l e d s u b j e c t s w ith lim b a m p u t a t i o n I. A g r e e m e n t sta tis tic s b e t w e e n M U A C a nd B M I w e r e a s f o l l o w s : k = 0 . 7 1 7 ( in c o n t r o l s u b je c ts).
M l A C > , ' M c m . M l A C • 24..* c m . ' HMI •. I X. 5.
B M I ' 1N.5.
subjects and illc subjects with limb amputation (either by BM1E or BMI,,).
1 lie sensitivity, specificity, positive predictive value, negative predictiv e value, and efficiency of BMI() and BMIE are shown in Table 4. To test the validity of BMI for assessment of nutritional stains. MI AC w as used as the gold standard. The sensitivity of BMI,, was = l()f; higher than that of BMIE, whereas for speci
ficity. the result was reversed. Although the sensitivity of BMIE was lower than that of BMIt). BMIf. had a much higher positive predictive value than did however, negative predictive value did not differ significantly between BMI0 and BMIE. The value of the efficiency statistic for BMIF tended to be higher than that of the efficiency statistic for BMI0 , but this difference was not significant. The efficiency of BMIE in the subjects with limb amputation was not significantly different from that of BMI in the control subjects. The binomial test forequality of proportions of sensitivity, specificity, and efficiency between MUAC and BMI in the control subjects and between MUAC and both BMI0
and BMIe in the subjects w'ith limb amputation (pooled) did not show any significant differences.
DISCUSSION
The aim of this study was to develop a method for estimating the total body weight of a person with limb amputation from current (postamputation) body weight by using anthropometric measurements and body weight proportions o f different limb segments according to Osterkamp (16). This method will ulti
mately help in assessing the nutritional status of persons with limb amputation. In the present study, the method was applied to an empirical data set collected from subjects with unilateral am
putation from Calcutta and adjoining areas, and the validity of the method was tested statistically.
It has been argued that stature is an important measurement for evaluating nutritional status. However, although measurement of stature is possible for persons with unilateral amputation who
TABI.K 4
Validity ot HMI c a l c u la t e d fr o m o b se r v e d b o d y w e i g h t ( B M I 0 ) a n d o f B M I c a l c u la t e d f r o m e s tim a ted b o d y w e i g h t (B M 1 E) fo r a s s e s s m e n t o f nutritional status in sub jects w it h li m b a m p utation and in c o n t r o l s u b je c ts w ith the u s e o f m id u p p e r arm c ir c u m f e r e n c e ( M U A C ) c la s s if ic a t io n as t h e g o l d standard
S e n s it iv it y S p e c if i c it y P o s i t i v e p r e d ic t iv e v a lu e N e g a t i v e p r e d ic t iv e v a lu e E f f ic ie n c ; M L A C co m p a re d w it h B M I , ,
B e lo w knee 7 7 . 7 8 8 6 . 5 4 6 6 . 6 7 9 1 . 8 4 8 4 . 2 9
A b o v e knee 1 0 0 .0 0 8 7 . 1 0 2 0 . 0 0 1 0 0 . 0 0 8 7 . 5 0
Pooled 7 8 . 9 5 ' 8 6 . 7 5 ' 5 7 . 6 9 9 4 . 7 4 8 5 . 2 9 '
M U A C c o m p a r e d w it h B M I E
B e l o w knee 6 6 . 6 7 9 4 .2 3 8 0 . 0 0 8 9 . 0 9 8 7 . 1 4
A b o v e k nee 1 0 0 . 0 0 1 0 0 . 0 0 1 0 0 . 0 0 1 0 0 . 0 0 1 0 0 . 0 0
P o o le d 6 8 . 4 2 ' 9 6 . 3 9 ' 8 1 . 2 5 9 3 . 0 2 9 1 . 1 8 '
M U A C c o m p a r e d w it h B M I
C ontrol s u b je c ts 7 8 . 9 5 9 4 . 1 9 7 5 . 0 0 9 5 . 2 9 9 1 . 4 3
' \ o t s i g n i f i c a n t l y d if f e r e n t f r o m t h e c o r r e s p o n d i n g v a l u e in t h e c o n t r o l s u b j e c t s ( b in o m ia l test for e q u a lit y ) .
wear a prosthesis, it is difficult for persons with bilateral ampu
tation. In the present study, stature measurements of the subjects with unilateral amputation were cross-checked by using the ratio of sitting height to stature, and these ratios were compared with those in healthy subjects without limb amputation. The ratio of sitting height to stature observed in the population in the present study corroborates the value found by Drillis and Contini (31).
As shown in Table 2, estimated weight was higher than ob
served weight because estimated weight compensates for the weight loss due to amputation. Testing the validity of the estimate was not possible because no such test, especially for living sub
jects, exists; however, the logic by which the method was devel
oped is in agreement with that of Himes (17).
For assessing the nutritional status of the subjects with limb amputation, BMI was calculated. In the present study, BMI was calculated from both the W0 and the WE of the subjects with limb amputation. The BMI calculated from W0 was less than that calculated from WE. BMI values in subjects with limb amputa
tion are less than those in healthy control subjects without limb amputation because the lost weight of the limbs i s not considered in calculating BMI (22). Therefore, to reduce the underestima
tion of nutritional status in persons with limb amputation, esti
mation of body weight is necessary so that BMI can be reliably estimated for persons with limb amputation. In the present study, a method for estimating body weight was developed, and its validity was tested. Estimated body weight was then used to reliably estimate BMI in subjects with limb amputation.
With classification of the subjects on the basis of MUAC and either BMI0 or BMIE (Table 3), there was a fair chance of un
derestimating the nutritional status (ie, CED or normal) of sub
jects with unilateral amputation if BMI0 was used for screening because of the weight loss due to limb amputation. Screening with BMIq also mismatched the classification based on MUAC.
However, if We was used for calculating BMIE, then the chances of underestimation were eliminated, and the BMI screening for CED and normal nutritional status corresponded with the clas
sification based on MUAC. The agreement statistics showed a stronger association between MUAC and BMIE than between MUAC and BMIq. In addition, the agreement statistics showed a stronger association between MUAC and BMI in the control subjects and between MUAC and BMIE in the subjects with limb amputation than between MUAC and BMI0 in the subjects with limb amputation.
The validity of BMI0 and BMIE were tested by calculating the statistics of sensitivity, specificity, positive predictive value, negative predictive value, and efficiency by using MUAC as the gold standard. Although the sensitivity of BMI0 was higher than that of BMIb, the specificity of BMIE was higher, and the highest value for specificity of BMIE was for the subjects with amputa
tion above the knee (perhaps because of the small sample size).
Furthermore, the positive predictive value of BMIE was higher than ttaat of BMIq, whereas negative predictive value did not differ significantly between BMIq and BMIE. However, the ef
ficiency of BM1e was higher than that of BMI0 . Binomial tests for equality of proportions failed to show any significant differ
ences in sensitivity, specificity, or efficiency between BMI0 in the subjects with limb amputation and BMI in the control sub
jects or between BMIE in die subjects with limb amputation and BMI in the control subjects.
It p an , the agreement statistics suggest that BMIE is a better ritimUnrofBMl than is BMIq for persons with limb amputation
and that nutritional status can be assessed on the basis of BMIe.
However, the specificity of the estimation is greater than its sensitivity for nutritional assessment.
The present study had some limitations, and thus some further research is required. The proportional weight estimation of the stump was performed by considering the length proportion al
though the compositions of the different segments of the body are not uniform. Therefore, a more elaborate study considering the three-dimensional structure and composition of different limb segments is necessary. For persons with bilateral amputations, accurate measurement of stature is difficult. However, the lengths of other segments of the body are generally estimated from stature. Therefore, population-specific data are required to calculate ratios of different body segments. Finally, further stud
ies are necessary to cross-validate the present estimating method.
Although the method of weight estimation may be valid for persons with bilateral amputations, because of the absence of empirical data, the validity o f the method could not be assessed;
thus, empirical studies are required.
W e are grateful to the subjects w h o participated in this study for their kind h e lp an d cooperatio n. W e thank Partha P Majumder, w h o provid ed many u se fu l su g g e stio n s and review ed the manuscript.
B o t h authors participated in the study d e s ig n , data analysis, and the writing o f t h e m anuscrip t. A M co llected the f ie ld data for the present study. Neither a u thor had a n y financial o r personal con flicts o f interest in the organization that supported the research.
REFERENCES
1. Braune W , Fisc her O. On th e center o f gravity o f the h um an body as related to the e quip m ent o f th e G erm an infantry soldier. M a q u e t PGJ.
Fu rlong R, trans. Berlin: Sprin ger-V er la g, 1985.
2 . D e m p st e r W T. Space requir em ents o f the seated operator. Wright- Patterson A ir Force B a se, O H: U S A ir Force. 1955. [ W A D C Technical R ep 5 5 -1 5 9 .]
3. C la u ser CE, M c C o n v ille JT, Y o u n g JW. W eight, vo lu m e and center of m a ss o f segm en ts o f the hum an b o dy . Wright-Patterson A ir F orce Base.
OH: U S A ir Force, 1969. [ A M R L - T R - 6 9 - 7 0 .]
4 . H a n a v a n EP. A mathem atical m o d e l o f the human b ody. Wright- Patterson Air Force B a se, OH : U S A ir Force. 19 64. [AMRL-TR-64*
102.]
5. B ernstein N A , S alzgeber O A , P a v le n k o PP, Gurvich N A . Determination o f lo c a tio n o f the centers o f gravity an d m ass o f the li m bs o f the living h um a n b o d y . M o s c o w : A ll U n i o n Institute o f Exp erimen tal Medicine,
1 9 3 6 (in Russian).
6. D rillis R , Contini R, B lu estein M . B o d y seg m en t parameters: a survey of m ea su rem en t techniques. A r t if L im b s 1 9 6 4 ; 8 : 4 4 - 6 8 .
7 . Jen sen R K . B o d y segm ent m a s s, radius and radius o f gyration propor
tio n s o f children. J B io m e ch 1 9 8 6 ; 1 9 : 3 5 9 - 6 8 .
8. Forbes G B . B o d y com position: in f lu e n c e o f nutrition, d isea se, growth and ag in g. In: Shils ME, O ls o n JA , S h ik e M . eds. M odern nutrition in health an d disease. Philadelphia: L e a & Febiger, 1 9 9 4 : 7 8 1 - 8 0 1 . 9. A u s m a n L M , Russel R M . N utritio n in the elderly. In: S h ils M E . Olson
JA , S h ik e M , eds. M odern nutrition in health and d is ease. Philadelphia:
L e a & Febiger, 1 9 9 4 : 7 7 0 - 8 0 .
10. T akam ura K, O hyam a S, Y a m a d a G , Ishinishi N . C h a n g es in body proportio ns o f Japanese m e d ic a l stud en ts b etw een 1961 and 1986. Ami P h y s A nthropol 1 9 8 8 ; 7 7 : 1 7 - 2 2 .
11. H a m il P V , Johnston FE , L e m e s h o w S . B o d y w eig h t, stature, and sitting height: w hite and N egro yo u ths, 1 2 - 1 7 years, United States. Rockville, M D : U S Dep artmen t o f H ealth E d u c a tio n and W elfare, 1973. [Publica
tion no. H R A -7 4 -1 6 0 8 .]
12. K au tz L , Harrison GG. C o m pa riso n o f b o d y proportions o f one-year-old M e x ica n -A m e ric a n and A n g l o children. A m J Public Health I98l;7k 2 8 0 - 2 .
13. W ilso n S R , Lo esc h D Z. Principal co m p o n e n t analysis o f shap e variable in adult individuals. Ann Hum B io l 1 9 8 2 ; 9 : 4 11 - 2 3 .
14. M artorell R, Manila RM, C a stillo R O , M e n d o za FS, P a w so n IG. Body
p r o p o r t io n s in t h r e e e t h n ic g ro u p s: c h i l d r e n a n d y o u t h s 2 - 1 7 y ea r s in N H A N E S II a n d N H A N E S . H u m B i o l 1 9 8 8 ; 6 0 : 2 0 5 - 2 2 .
15. T a n n e r J M . F o e t u s in to man: p h y s i c a l g r o w t h f r o m c o n c e p t i o n to m a turity. L o n d o n : O p e n B o o k s P u b lis h in g , 1 9 7 8 .
16. O s t e r k a m p L K . C urrent p e r s p e c t iv e o n a s s e s s m e n t o f h u m a n b o d y p r o po rtio n s o f r e l e v a n c e to a m p u te es . J A m D i e t A s s o c 1 9 9 5 ; 9 5 : 2 1 5 - 8 . 17. H im e s JH . N e w eq u a t io n to e s t im a t e b o d y m a s s in d e x in am p u te es. J A m
D ie t A s s o c 1 9 9 5 ; 9 5 : 6 4 6 .
18. T z a m a l o u k a s A H , Patron A , M a lh o tr a D . B o d y m a s s in d e x in am p u te es.
J P E N J P a r e n t e r E n tera l N u tr 1 9 9 4 ; 1 8 : 3 5 5 - 8 .
19. T / a m a l o u k a s A H , L e g e r A . H ill J, M u r a t a G H . B o d y m a s s in d e x in p atie n ts w i t h a m p u t a t i o n s o n p e r it o n e a l d ia ly s is : error o f u ncor re cted e s t im a t e s a n d p r o p o s e d co r r e c tio n . A d v P e r it D i a l 2 0 0 0 ; 1 6 : 1 3 8 - 4 2 . 20. D e m p s t e r W T , G a u g h r a n G R L . P r o p e r t ie s o f b o d y s e g m e n t s b a se d o n
siz e a n d w e i g h t . A m J A n a t 1 9 6 7 ; 1 2 0 : 3 3 - 5 4 .
21. S h e tty P S . J a m e s W P T . B o d y m a s s in d e x : a m e a s u r e o f c h ro n ic e n er g y d e f i c i e n c y in a d u lt. R o m e: F o o d an d A g r ic u l t u r a l O r g a n iza tio n , 1 9 9 4 .
| F A O F o o d a n d N u tr itio n p aper, n o . 5 6 . ]
22. M ad an M . P e l e s E , H a lk in H , e t al. I n c r e a s e d ca r d io v a sc u la r d is e a s e m o rta lity ra t e s in tra u m a tic l o w e r l i m b a m p u t e e s . A m J C a rd io l 1 9 98;
8 2 : 1 2 4 2 - 7 .
23. Lau ner LJ. H a rris T , R u m p e l C , M a d a n s J. B o d y m a s s in d e x , w e i g h t c h a n g e , a n d r isk o f m o b ilit y d is a b ility i n m i d d l e - a g e d and o ld er w o m e n . T h e e p i d e m i o l o g i c f o ll o w - u p s t u d y o f N H A N E S 1. J A M A 1 9 9 4 :2 7 1 ; 1 0 9 3 - 8 .
24. K w o k T . W o o J, L a u E. Pr ed iction o f b o d y f a t b y an thropom etry in o ld er C h i n e s e p e o p l e . O b e s R e s 2 0 0 1 ; 9 : 9 7 - 1 0 1 .
2 5 . F reem a n A C , Y o u s a f z a i A K , F ilte a u S M , P a i M . S e r u m le ptin in d is a b led a nd n o n - d i s a b l e d ch ildren in an I n d ia n s lu m p o p u la tio n . E ur J C lin N utr 2 0 0 2 ; 5 6 : 9 6 7 - 7 2 .
2 6 . W e in e r JS, L a u r ie J A . Practical h um an b i o lo g y . L o n d o n : A ca d e m ic Press, 19 8 1 .
2 7 . Institute for C o n s u m e r E r o g o n o m ic s . S eated a n th r o p o lo g y : the prob
le m s i n v o lv e d in a la r g e s c a le sur vey o f d is a b le d and elde rly people.
E r g o n o m ic s 1 9 8 1 ; 2 4 : 8 3 1 - 4 5 .
2 8 . G o s w a m i A , G a n g u li S , C ha terjee B B . A n th r o p o m e tr ic characteristics o f d isabled and n o r m a l India n m en . E r g o n o m ic s 1 9 8 7 ; 3 0 : 8 1 7 - 2 3 . 2 9 . N o w a k E . W o r k s p a c e f o r d is a b l e d p e o p le . E r g o n o m i c s 1 9 8 9 ;3 2 :
1 0 7 7 - 8 8 .
3 0 . M i t c h e ll C O , L ip s c h i t z D A . A r m le n g th m e a s u r e m e n t as an alternative t o h eig h t in n utr itio n a l a s s e s s m e n t o f th e elderly . J P E N J Parenter Enteral N utr 1 9 8 2 ; 6 : 2 2 6 - 9 .
3 1 . D rillis R , C o n t in i R . B o d y s e g m e n t param eters. N e w Y ork: O f f ic e o f V o c a tio n a l R e h a b ilit a t io n , D e p a rtm en t o f H e a lth , E d u c a t io n and W e l
fare, 1 9 6 6 . [ R ep o rt n o . 1 1 6 6 - 0 3 . ] C ited by: O b o m e D J , ed . Erg o n o m ic s at work. N e w Y o rk : J o h n W i l e y and S o n s , L td, 1 9 8 6 : 4 3 .
3 2 . M o z u m d a r A , R o y S R . A n o t e o n the alternativ e m e t h o d for the a s se ss m e n t o f nutritional sta tus. J Ind ia n A nth rop ol S o c 2 0 0 0 ; 3 5 : 1 9 7 - 2 0 3 . 3 3 . J a m es W P T , M a s c i e - T a y l o r C G , N o r g a n N G , B is tr a in B R , Shetty PS,
Ferro-L u zzi A . T h e v a lu e o f arm c i r cu m fer e n c e m ea s u r e m e n ts in as
s e s s i n g c h r o n ic e n e r g y d e f ic ie n c y in Third W o rld ad ults. Eur J C lin Nutr 1 9 9 4 ; 4 8 : 8 8 3 - 9 4 .
3 4 . L a n g T A , M i c h e l l e S . H o w to report statistics in m e d ic in e . Philadelphia:
A m e r ic a n C o lle g e o f P h y s ic i a n s , 1997.
3 5 . M e d e c in s sans F r o n tier es . G u id e lin e s o n the su p p lem e n ta r y feeding program s. B r u sse ls: M S F , 199 3 .
3 6 . Y o u n g H. F o o d s c a r c it y a nd fam in e: a s s e s s m e n t a n d re sp o n se . Oxford, U n ited K in g d o m : O x f a m , 199 2 .
3 7 . B riend A , G a r e m e M , M a i r e B , Fon tain e O , D i e n g K . N utr ition al status, a g e and survival: th e m u s c l e m a s s h y po th esis. E u r J C lin N u tr 1989;43:
7 1 5 - 2 6 .
