© Springer Science + Business Media, Inc. Obesity Surgery, 17, 2007 565
Obesity Surgery, 17, 565-568
Weight loss, although certainly not the only criteri-
on, is an important outcome measure after bariatric
surgery. However, as the years go by and health con-
ditions in society change, it is important to review
the current situation.
In 1959, the Metropolitan Life Insurance Company
released tables of the best weight for each height for
longevity, based on collected insurance data.1These
were called the “desirable” weights.
In 1983, the Metropolitan Life Insurance
Company in New York released the “ideal” weights
for greatest longevity (or for lowest mortality).2
These were based on the Build Study of 1979, deter-
mined by the Society of Actuaries in Chicago, on
patients followed for 18 years (1954-1972).3This
data was collected from 25 life insurance companies
in the USA and Canada, representing 4.2 million
insured individuals, and is still by far the largest
study available. The “ideal” weights were higher
than the prior “desirable” weights, and this was
attributed to an increase in muscle mass due to
improved fitness.2The average weights in the popu-
lation are higher than the ideal weights for survival.
The criticisms of the 1983 Height and Weight
Tables were:41) the insured population tended to be
affluent and white; 2) 10% of the weights were self-
reported; 3) the weights included indoor clothing,
estimating 5 lbs for men and 3 lbs for women, and
shoes with 1 inch heels in both men and women; 4)
individuals with known heart attacks, cancer, dia-
betes and hypertension were excluded.
The ideal weight was based on the mean weight of
the insured individuals age 25-59, but ideal weight
for survival actually increased up to age 54.3
The Metropolitan Tables included small, medium
and large frames, based on elbow-girth using
calipers,5because the elbows do not develop adi-
posity. The elbow-girth from the 15th to 85th per-
centile was termed the medium frame.
The Tables presented weight ranges for height,
sex and body frame, associated with the lowest mor-
tality. The mid-point of the ideal weight range for
the medium frame for each height was selected as
the ideal weight for calculations.
Excess weight (initial weight – “ideal” weight) was
the weight above “ideal” weight. From this, the wide-
ly used equation became available: percent excess
weight loss (using the initial excess weight) or
%EWL =preoperative weight – current weight x100.
preoperative weight – ideal weight
An easy formula to calculate the “ideal” weight is
shown in Table 1.6
%EWL has been widely used. However, this is
based on 1979 data. People are now living longer.
The ideal weight associated with maximal longevi-
ty has increased,7likely due to earlier diagnoses,
medications, improvements in public health and
medical care, and the effect of the fitness industry
on muscle mass.
However, this current finding of increased longevi-
ty at higher weights may in the future become negat-
ed due to the occurrence today of obesity and its relat-
ed morbidities in adolescents and teenagers,8where
we now have a generation which may not live as long
as its parents.9In any event, the %EWL, widely used
in the USA and many other countries by bariatric sur-
geons, has lost considerable accuracy by 2007.
Reporting Weight Loss 2007
Table 1. Formula for calculation of Ideal Weight6*
Adult Female:5 ft.tall = 119 lb.For each additional inch, add 3 lb.
Adult Male: 5 ft. 3 inches tall = 135 lb. For each additional inch,
add 3 lb.
1 foot = 30.4 cm; 1 inch = 2.54 cm.
Divide lb. by 2.2 to change to kg.
*Formula corresponds to mid-point of medium frame of the
Metropolitan Tables, with accuracy within 1%. To convert to
ideal weight for small or large frame, decrease or increase the
result by 10%. Patients without shoes.
OS 17(5)-Articles 4/11/07 1:51 PM Page 565
The Society of Actuaries (SOA) does not plan to
perform an updated study, and indeed weight infor-
mation is no longer collected by some insurance com-
panies (personal communication, Jack Luff, SOA,
Chicago, July 12, 2005). Also, the Metropolitan Life
Insurance Company is not aware of any future update
of the Metropolitan Height and Weight Tables (per-
sonal communication, Brittany L. Eklebury, MPH,
MetLife Wellness & Fitness Services, New York, July
13, 2005). As such, %EWL has less relevance for the
bariatric surgical community.
Percent of Initial Weight Lost
Percent of the initial weight lost has been used in
some medical papers. It is problematic in that it
does not relate to a given height or degree of “fat-
ness”, which is germane to any weight and popula-
Body Mass Index (BMI)
Studies were made, starting in the early 20th
Century, for an index of relative body weight – i.e.
body weight in proportion to some function of
height, to provide an index number of body “fat-
ness”. Studies were made with W/H, W/H2
(Quetelet’s index) and W/H3(with W = Weight in
kg and H = Height in m). It was found that W/H2
gave the most accurate index when compared to
body hydro-densitometry studies.10Actually, for
men the optimal index was W/H2and for women
W/H1.5(women have an increased proportion of
body fat),11but W/H2has been used for both sexes
because of its simplicity. Comparing W/H2 calcula-
tion (“body mass index” or BMI) to the
Metropolitan Tables for “ideal” weight, and
researching the optimal BMI to avoid obesity-relat-
ed diseases, the range of 20-24.9 kg/m2was found.12
It was found that there was increased risk of mortal-
ity with BMI <18.5 kg/m2, and that BMI <17.5 sug-
gested anorexia nervosa13or cachexia. BMI of 18.5-
20 may be suspicious of an eating disorder.
The relationship between BMI and mortality fol-
lows a J-shaped curve.14,15Overweight is designated
as BMI 25-29.9, class I obesity 30-34.9, class II obe-
sity 35-39.9, class III (or morbid) obesity ≥40, super-
obesity 50-60, and super-super obesity ≥60 kg/m2.
Studies of 20,000 to 1,000,000 individuals have
determined optimal adult BMIs in the range of 20-
25,16,17but this number differs slightly for the onset
of the individual diseases of the metabolic syn-
drome. For example, insulin sensitivity for glucose
disposal was found to be impaired in normal volun-
teers when a BMI of 26.8 kg/m2was reached.18
BMI is being used widely for different popula-
tions, has good accuracy, and is easy to calculate. It
has been found, however, that Asians may already
develop impaired glucose tolerance and elevated
blood pressure at a lower BMI, e.g. 23.19-22This has
been attributed to an increased percentage of the
BMI made up of visceral adipose tissue in Asians.
Percent of Excess BMI Lost (%EBL)
The percent of excess BMI lost (percent EBMIL or
simply %EBL), is currently the best method for
comparing various obesity treatments, and is
becoming more widely used.17It is calculated from
%EBL =preoperative BMI – current BMI x 100.
preoperative BMI – 25
% EBL uses BMI 25 as the upper limit of normal,16
and is not quite the same as %EWL (based on the
Metropolitan Tables) which uses the mid-point of
the medium frame as their “ideal” weight. %EBL is
now being adopted as the standard measure for stud-
ies of many granting bodies.
Kg has been proposed, and used in a number of med-
ical studies, e.g. comparing the effect of drugs.17
However, whereas in a medical study, a loss of 7 kg
may be meaningful, this is generally not the goal in
bariatric operations on morbidly obese patients. Also,
kg by itself is not related to height, which varies con-
siderably in populations, and the healthy weight in a
tall subject will be higher than in a short subject.
Because body fat distribution studies have repeated-
ly confirmed that increased visceral (truncal) obesi-
ty is associated with the serious diseases of the
Deitel et al
566 Obesity Surgery, 17, 2007
OS 17(5)-Articles 4/11/07 1:51 PM Page 566
metabolic syndrome, there has been considerable
interest in adipose measurements (whose relevance
is also affected by physical activity, diet and hered-
ity).23The primary aim of bariatric surgery is to
decrease fat mass.
Waist circumference (WC) and waist-to-hip
ratio (WHR) have been proposed.14,24-26WC is
measured midway between the lateral lowest rib
margin and the iliac crest with the patient stand-
ing. The hips are measured at their largest circum-
ference over the greater trochanters. Healthy WC
in males has been stated to be ≤98 cm and in
females ≤88 cm. These values have not been vali-
dated for different heights. Simply stated, normal
WHR in males is <1.0 and in females ≤0.8. A high
WHR in a female is designated an “apple”
(android) build, whereas the lower level in an
obese female is designated “pear” (gynoid) build.
These precise measures vary with the authors or
study. However, increased WC has major implica-
tions for aspects of the metabolic syndrome, and is
associated with an increased incidence of type 2
diabetes, hypertension, dyslipidemia and cardiac
disease, and is reflected in nonalcoholic fatty liver
disease. WHR >.83 in the female and >.9 in the
male has been associated with a three-fold
increase in risk for myocardial infarction.27
However, again, true severity of the measurement
cannot be designated without a height.
White et al28found that BMI correlated more with
elevated blood pressure than did WHR and the sum
of skinfold measurements. Furthermore, BMI has
greater simplicity and precision than measuring
Percentage of fat mass (%FM) can be calculated
from the BMI by the formula of Gallagher et al,30
but the result is inaccurate in morbidly obese
patients.29Direct measurement of body composition
has also been used for changes in %FM, by dual-
energy x-ray absorptiomety (DEXA) and by bio-
electric impedance analysis (BIA). DEXA has been
confirmed to provide accurate data, while BIA was
generally inaccurate.29However, DEXA is a costly
study, based on absorption of x-rays. Fat measure-
ment by ultrasound has been found to be as accurate
as CT scanning, is less costly, and does not subject
the individual to radiation.31
Methods used to report weight loss in the bariatric sur-
gical literature have been multiple and occasionally con-
tentious. There will be further changes as the science
develops. OBESITY SURGERY prefers BMI, and for
comparison of weight loss within a study or between
studies, this Journal strongly encourages percent of
excess BMI loss (%EBL). However, %EWL (based on
the old Metropolitan Tables) has been used in many past
studies, so that this measure may still be accepted to
enable comparison. Body weight in kilograms can add
to the description, but has no value by itself, because kg
alone gives no indication of relative height.
With the effect of visceral obesity on the serious
diseases of the metabolic syndrome, waist circum-
ference or waist-to-hip ratio is becoming an impor-
tant measurement in certain related studies. In some
instances, more complex methods for measuring fat
distribution may be relevant.
Mervyn Deitel, MD, Editor-in-Chief
Khaled Gawdat, MD
Assistant Professor of Surgery
Ain-Shams School of Medicine
John Melissas, MD
President of IFSO
Professor of Surgery
Heraklion, Crete, Greece
The authors thank Dr. John B. Dixon for his input.
1. Metropolitan Life Insurance Company: New weight
standards for men and women. Stat Bull Metrop Life
Insur Co 1959; 40: 1-10.
2. 1983 Metropolitan Height and Weight Tables. New
York: Metropolitan Life Foundation. Statistical
Bulletin 1983; 64(1): 2-9.
3. Build Study, 1979. Society of Actuaries and
Association of Life Insurance Medical Directors of
America. Philadelphia: Recording and Statistical
4. Deitel M, Shahi B. Morbid obesity: Selection of
patients for surgery. J Am Coll Nutr 1992; 11: 457-62.
5. Frisancho AR, Flegel PN. Elbow breadth as a meas-
Reporting Weight Loss 2007
Obesity Surgery, 17, 2007 567
OS 17(5)-Articles 4/11/07 1:51 PM Page 567
ure of frame size for US males and females. Am J
Clin Nutr 1983; 37: 311-4.
6. Miller MA. A calculated method for determination of
ideal body weight. Nutritional support services 1985;
7. Flegal KM, Graubard BI, Williamson DF et al. Excess
deaths associated with underweight, overweight and
obesity. JAMA 2005; 293: 1861-7.
8. Bhargava SK, Sachdev HS, Fall CH et al. Relation of
serial changes in childhood body-mass index to
impaired glucose tolerance in young adulthood. N
Engl J Med 2004; 350: 865-75.
9. Olshansky JF, Passero DJ, Hershow RC et al. A
potential decline in life expectancy in the United
States in the 21st Century. N Engl J Med 2005; 352:
10.Keys A, Flaminio F, Kervonen MJ et al. Indices of rel-
ative weight and obesity. J Chron Dis 1972; 25: 329-43.
11.Abraham S, Carroll MD, Najjar MF et al. Obese and
overweight adults in the United States (Vital and
Health Statistics ser 11, no 230) (DHHS publ no
[PHS] 83-1680), National Center for Health Statistics,
Hyattsville, MD, 1983.
12.Calle RR, Thun MJ, Petrelli JM et al. Body-mass
index and mortality in a prospective cohort of U.S.
adults. N Engl J Med 1999; 341: 1095-105.
13.Deitel M. Anorexia nervosa following bariatric sur-
gery. Obes Surg 2002; 12: 729-30.
14.Lemieux S, Mongeau L, Paquette M-C et al. Health
Canada’s new guidelines for body weight classifica-
tion in adults: challenges and concerns. CMAJ 2004;
15.MacLean LD. Surgery for obesity: where do we go
from here? Am Coll Surg Bull 1989; 74: 20-3.
16.Kuczmarski RJ, Flegal KM. Criteria for definition of
overweight in transition: background and recommen-
dations for the United States. Am J Clin Nutr 2000;
17.Dixon JB, McPhail T, O’Brien PE. Minimal reporting
requirements for weight loss: current methods not
ideal. Obes Surg 2005; 15: 1034-9.
18.Campbell PJ, Gerich JE. Impact of obesity on insulin
action in volunteers with normal glucose tolerance:
demonstration of a threshold for the adverse effect of
obesity. J Clin Enodcrinol Metab 1990; 70: 114-8.
19.Deurenberg-Yap M, Schmidt G, van Staveren WA et
al. The paradox of low body mass index and high
body fat percentage among Chinese, Malays and
Indians in Singapore. Int J Obes 2000; 24: 1011-7.
20.Choo V. WHO reassesses appropriate body-mass
index for Asian populations. Lancet 2002; 360: 235.
21.Ko GTC, Tang J, Chan JCN et al. Lower body mass
index cut-off value to define obesity in Hong Kong
Chinese: an analysis based on body fat assessment by
bioelectrical impedance. Br J Nutr 2001; 85: 239-42.
22.Jafar TH, Chaturvedi N, Pappas G. Prevalence of
overweight and obesity and their association with
hypertension and diabetes mellitus in an Indo-Asian
population. CMAJ 2006; 175: 1071-7.
23.Lee W-J, Wang W, Chen T-C et al. Clinical signifi-
cance of central obesity in laparoscopic bariatric sur-
gery. Obes Surg 2003; 13: 921-5.
24.Janssen I, Katzmarzyk PT, Ross R. Waist circumfer-
ence and not body mass index explains obesity-relat-
ed health risk. Am J Clin Nutr 2004; 79: 379-84.
25.Poirier P, Lemieux I, Mauriege P et al. Impact of waist
circumference on the relationship between blood
pressure and insulin: the Quebec health survey.
Hypertension 2005; 45: 347-8.
26.Merchant A, Yusuf S, Sharma AM. A cardiologist’s
guide to waist management. Heart 2006; 92: 865-6.
27.Yusuf S, Hawken S, Ounpuu S et al. Obesity and the
risk of myocardial infarction in 27,000 paticipants
from 52 countries: a case-control study. Lancet 2005;
28.White FM, Pereira LH, Garner JB. Associations of
body-mass index and waist:hip ratio with hyperten-
sion. CMAJ 1986; 135: 313-20.
29.Alvarez VP, Dixon JB, Strauss BJB et al. Single fre-
quency bioelectrical impedance is a poor method for
determining fat mass in moderately obese women.
Obes Surg 2007; 17: 211-21.
30.Gallagher D, Heymsfield SB, Heo M et al. Healthy
percentage body fat ranges: an approach for develop-
ing guidelines based on body mass index. Am J Clin
Nutr 1990; 51: 571-7.
31.Pontiroli AE, Pizzocri PI, Giacomelli M et al.
Ultrasound measurement of visceral and subcuta-
neous fat in morbidly obese patients before and after
laparoscopic adjustable gastric banding: comparison
with computerized tomography and with anthropo-
metric measurements. Obes Surg 2002; 12: 648-51.
Deitel et al
568 Obesity Surgery, 17, 2007
OS 17(5)-Articles 4/11/07 1:51 PM Page 568