This item was submitted to Loughborough’s Institutional Repository

(https://dspace.lboro.ac.uk/) by the author and is made available under the

following Creative Commons Licence conditions.

For the full text of this licence, please go to:

http://creativecommons.org/licenses/by-nc-nd/2.5/
Bulletin der Schweizerischen Gesellschaft für Anthropologie 18(2): 5–11 (2012)
Bulletin de la Société Suisse d’Anthropologie 18(2): 5–11 (2012) ISSN 14208 - 4835

The Body Mass Index: the Good, the Bad, and the Horrid
BARRY BOGIN AND INES VARELA-SILVA

Centre for Global Health and Human Development, School of Sport, Exercise and Health Sciences, Loughborough University, United Kingdom

Summary
The Body Mass Index (BMI) was developed to estimate the risk for overweight in large samples of people from the wealthy, heavily
industrialized nations of Western Europe and North America. When used for this purpose the BMI is, generally, a good tool to estimate
overweight. The BMI is a bad tool when used to estimate fatness prior to the onset of the obesity epidemic in 1980 because BMI
cannot distinguish between fat and lean tissue and there was, generally, lower levels of fatness in the general population before that
date. The BMI is also a bad tool when used to estimate fatness for individuals in any nation or in any group of people. The BMI was
never intended to be used for individual diagnosis. The BMI becomes a horrid tool to estimate fatness or health risk when used in
some groups of people, such as high-level athletes, body building enthusiasts, people engaged in jobs with strenuous physical activity,
and in groups suffering from the nutritional double-burden of very short stature with high body fatness.

Keywords: BMI, body composition, nutritional dual-burden, Maya

Introduction measured in kilograms was proportional to the square of

the height measured in meters (Quetelet 1832). This
There was a little girl, ratio was given the name Quetelet Index (QI). By the
Who had a little curl, mid-20th Century the QI or other related weight-for-
Right in the middle of her forehead.
height ratios were used by some human biology
When she was good,
She was very good indeed,
researchers to assess fatness and by the life insurance
But when she was bad she was horrid.
industry to apportion risk and insurance premiums
(Dublin et al. 1937, Billewicz et al. 1962, Khosla and
By Henry Wadsworth Longfellow, 1807–1882 Lowe 1967).
Eknoyan (2008: p 48) reviews the use of Quetelet’s
The Body Mass Index, abbreviated BMI, is a widely Index and reports that, “One of the first studies to
used ratio of weight-for-height. BMI is calculated as confirm the validity of the Quetelet Index in
[weight in kilograms / (height in meters)2]. Similar to the epidemiological studies comprised data gathered during
subject of Longfellow’s poem, when BMI is used the fourth examination of the Framingham study [Florey
appropriately it is ‘very good indeed.’ When BMI is 1970]. In a subsequent comparative study of available
used inappropriately the consequences range from ‘bad indices of relative weight and obesity published in 1972,
to horrid’. Ancel Keys (1904–2004) confirmed the validity of the
Quetelet Index and named it the Body Mass Index
(BMI) …” (Keys et al. 1972). The importance of the
The ‘Good’ of BMI Framingham Study must be stressed. It was one of the
first well-designed epidemiological investigations of the
Lambert Adolphe Quetelet (1796–1874) published causes of heart disease, which was then, as now, a major
the first statistically complete studies of the growth in public health concern. The prominence of Ancel Keys in
height and weight of children. Quetelet was the first nutritional science and public health policy must also be
researcher to make use of the concept of the “normal stressed. Together, the fear of heart disease and the
curve” (commonly called today the normal distribution reputation of Keys elevated the BMI to international
or “bell-shaped” curve) to describe the distribution of prominence.
his growth measurements, and he also emphasized the Higher BMI scores indicate that an individual has
importance of measuring samples of children, rather relatively more weight-for-height than a person with a
than individuals, to assess normal variation in growth. In lower score. The value of BMI indicates only this and
1832 Quetelet proposed that normal body weight does not provide any information about body
 B. BOGIN & I. VARELA-SILVA

composition, that is, relative amounts of lean tissue men are classified by BMI as having excess adiposity
versus fat tissue. Even so, in the general population of when by more direct measures of %body fat they are
the wealthier nations of Europe, North America, within the desirable range, while others have ‘normal’
Australia, New Zealand, and Japan a higher BMI score BMI with excess fatness.
usually indicates more body fatness. This is due to the Despite the inappropriateness of BMI for use with
lack of physical activity and, often, excessive food the individual person it continues to be used in this way
consumption of the majority of people in these nations. quite commonly in research and clinical practice. One
In special groups within the population, such as highly researcher/practitioner writes, “The best documented
trained athletes or body building enthusiasts, a higher measure of obesity is the body mass index (BMI)…
BMI score may be due to increased muscle mass. In the which is now used almost universally in adults and
middle and low income nations of Africa, Asia, and the increasingly in children” (Cole 2003). It is easy to find
Americas the transitions in diet and labor patterns which online BMI calculators which allow people to discover
emulate those of the rich nations are bringing about their BMI and use this value to diagnose their risk for
greater fatness of the population. In general, BMI serves heart disease, diabetes, anorexia nervosa and other
well to assess this rising tide of fatness, overweight and ailments without any professional medical supervision.
obesity. BMI is everywhere. A search of PubMed using the term
‘BMI’ finds 60,840 articles dating back to 1978. A
search using the terms ‘BMI + fatness’ scored 21,588
results, the first on the list being titled, Body Mass Index
The ‘Bad’ of BMI and Calculator: Understand Calorie Count of Important
Foods and Keep Meals Under 300 Calories by Steve
This application of BMI may be useful to assess
Ryder. Assuming that Ryder’s use of ‘Calories’ mean
relative fatness for large groups of people, but the BMI
kilocalories, the first author of this paper (BB) would
of individual people should not be interpreted in this
need to eat nine such 300 kcal meals a day to meet his
manner. Quetelet never intended that his Index be used
energy needs.
for individuals – he developed the index for large
A good deal of skepticism in the BMI was generated
samples so that he could construct and interpret the
by the work of Stanley Garn (1922–2007). Garn was our
distribution of height-for-weight along normal curves.
colleague at the University of Michigan and by the late
Keys et al. (1972) warned that the BMI should not be
1970s he was voicing concern about the misuse of the
used for individual diagnosis due to complex effects of
BMI. A few years later he published the article “Three
age and sex in the mathematical determination of the
limitations of the body mass index” (Garn et al. 1986).
value of the BMI and the poor precision of that value to
The three limitations of BMI to assess fatness are: 1)
predict health problems of an individual. Today we
BMI is not independent from stature. As a ratio, the
know that ethnicity also has important effects on the
calculation of BMI should yield the same result for all
determination of BMI values, desirable weight, and
combinations of identical weight-for-height. Garn et al
fatness (Razak et al. 2007). Indeed, sex, age, and
showed this is not true as there is a change in the
ethnicity all interact to further confound the meaning of
correlation between stature and the BMI from about
individual BMI values.
+0.30 for children to an average of -0.12 for women
There are many critiques of the misuse of BMI when
20–39 years old; 2) people of the same height have
applied to individuals. Ross and Eiben (2002: p. 49),
different BMI values according to frame size and
“…draw attention to some embarrassing evidence…” in
relative leg length. People with narrower chests and/or
the literature showing that a common BMI scale for men
longer legs relative to their total height have lower BMI
and women is a mathematical artifact which does not
values; 3) the BMI cannot distinguish between the
relate to empirical biology, that BMI may not predict the
amount of lean tissue and fat tissue of a person’s body.
sum of skinfolds much better than chance, and that BMI
This is not only a problem for athletes versus sedentary
cannot distinguish fat from muscle. A graph available
people of the same height and weight but also is part of
online at http://en.wikipedia.org/wiki/File:Correlation_
the cause for ethnic, age and sex effects on the BMI.
between_BMI_and_Percent_Body_Fat_for_Men_in_N
All three of the limitations highlighted by Garn and
CHS%27_NHANES_1994_Data.PNG, prepared by
colleagues are, in fact, interrelated. Greater stature may
Mark Warren and based on data from Romero-Corral et
be associated with a narrower skeletal frame, and frame
al. (2006), shows that use of the BMI mis-classifies the
size may be associated with total muscle mass. The
body fatness of 24.6% of adult men in the United States
effects of relative leg length have been especially well
measured for the National Health and Nutrition
studied and subsequent work confirms that when
Examination Survey, 1994 (sample size = 8550). Some

6 Bulletin de la Société Suisse d’Anthropologie 18 (2012)

 The Body Mass Index

matched for total stature, the people with relatively short study are based on the population of all infants born in
legs have higher BMI regardless of their percentage of England, Scotland, and Wales during March 3rd to 9th,
body fatness (Norgan and Jones 1995, Deurenberg and 1958. Lasker and Mascie-Taylor (1989) published the
Deurenberg-Yap, 2003, Bogin and Beydoun 2007, mean height, weight, and BMI of these boys and girls at
Bogin and Varela-Silva 2008). The study by Bogin and ages seven, 11, and 16 years stratified by the social class
Beydoun analyzed the data for adults 20.00–49.99 years of the male head of household. Lasker and Mascie-
old from the Third National Health and Nutrition Taylor (1996) also published mean heights at age 23
Examination Survey, 1988–1994. This is a nationally years for these same samples. There are two to three
representative population survey and includes men and thousand individuals in each age group. “In Britain,
women of European (White), African (Black) and social class is officially ascribed on the basis of the
Mexican ancestry. The relative leg length effect on the occupation of the male head of the household … the
BMI is statistically significant for both sexes and all Registrar General’s 5-fold class designations … are …
three ancestry/ethnic groups. It is important to note, social class I – professional; II –intermediate; III –
however, that in terms of statistical magnitude the most skilled; IV – semi-skilled; and V – unskilled …” (Lasker
important variable associated with the BMI of these and Mascie-Taylor 1989, p. 1). Lasker and Mascie-
people is the sum of four skinfolds (subscapular, triceps, Taylor find that mean stature and weight are
suprailiac, thigh). The ‘sum of skinfolds’ variable significantly related to social class, and decline,
explains about 74% and 87% of the variance in BMI generally, from social class I and II to V at each age
values in the different combinations of sex-ancestry (e.g. (Figure 1 A and B). Male or female sex is also a
Black women, Mexican men, etc.). This lends credence significant influence on height and weight, with girls
to the ‘good’ of BMI. The relative leg length variable and women being, on average, shorter and lighter than
explains an additional 4% of the variance in BMI, and boys and men. There was no SEX by SES interactions in
this hints at the ‘bad’ of BMI. any statistical analysis, and only the data for males are
Even though the statistical effect size of relative leg shown here. The statistical impact of the social class
length on BMI is small, compared with the effect of effect on stature and weight is achieved by age seven
fatness as measured by skinfolds, that effect is both years, and is then maintained through age 23 years.
statistically and biologically real and important. In the In contrast to the SES related pattern for height and
United States men and women with relatively shorter weight by social class over time, BMI follows a different
legs carry more subcutaneous fat, as measured by the trend. At age 7 and 11 years all social classes are about
sum of four skinfolds, than adults with relatively longer equal, with classes II and IV at bit higher than the others.
legs. Why this is so is not understood at present. Bogin Between ages 11 and 16 years there is a clear change in
and Beydoun (2007) offer a possible explanation based the pattern for mean BMI values. Lasker and Mascie-
on research in human life history biology. In brief, they Taylor report that the three lower social classes, III, IV,
propose that poor nutrition and health during the pre- and V, have greater increases in BMI compared with
natal, infancy and childhood stages of growth results in social classes I and II. The change in BMI by social
relatively shorter legs and a modified physiology which class, “… is due mostly to the taller stature of social
tends to store body fat when excess energy is available class I [and II] youths of both sexes, not to the obesity of
(Varela-Silva et al. 2007, Bogin and Varela-Silva 2010). those of [lower] social class ...” (p. 5). None of the social
No matter what the cause, the leg length effect, along classes is obese in the current usage of that word, so I
with sex, ethnicity, age, physical activity and many other believe the authors mean the fatness of the different
factors may help explain why the adiposity of nearly social classes when using the word ‘obesity.’ Indeed,
25% of men in the United States is mis-classified by none are ‘fat’, that is overweight, by current references
BMI. for BMI. The range of mean BMIs at age 16 years for all
social classes is 19.9 – 21.1, which are at or below the
50th percentile of BMI for 16 year olds in the United
States measured from 1971 to 1980 (Frisancho 1990).
The ‘Horrid’ of BMI BMIs in this range are considered ‘healthy.’
One question from these British data is, if greater
In the second edition of the book Patterns of Human
BMI is not measuring fatness then what is it measuring?
Growth (Bogin 1999) the BMI is mentioned only nine
A possibility is a hypothesis offered by Christian
times across 398 pages of text. The nine references are
Aßmann and Michael Hermanussen (in press) that there
all in relation to an analysis of The National Child
is a socially and psychologically influenced community-
Development Study of the Great Britain, a longitudinal
based target for height and other body size dimensions.
study of growth in height and weight. Data from this

Bulletin der Schweizerischen Gesellschaft für Anthropologie 18 (2012) 7

 B. BOGIN & I. VARELA-SILVA

Their hypothesis builds on empirical research that for their height. Stunted people may also have
Insulin-like Growth Factor-1 (IGF-1) levels in the blood disproportionately short legs and a relatively larger
are associated with social position. IGF-1 is a major trunk length for their height. Any or all of these effects
promoter of cellular growth and people with more IGF- of stunting may increase BMI without an associated
1 during the years of growth are generally taller. Kumari increase in body fat.
et al (2008) measured IGF-1 levels in the participants of Our research team includes members from the
the 1958 British Birth Cohort – the same sample Centro de Investigacion y de Estudios Avanzados del
analysed by Lasker and Mascie-Taylor (1989). Kumari Instituto Politecnico Nacional (Cinvestav) in Merida,
et al. looked for associations of IGF-1 with the social Mexico and Loughborough University in the United
position of the participants as measured by their father’s Kingdom. We are working with participants from the
or their own occupational class at three time points in Maya ethnic group of the Yucatan Peninsula. Our overall
childhood and adulthood. They found that low social research goal is to understand why the Maya people,
position is associated with lower levels of IGF-1. Other both children and adults, show high levels of stunting
research finds that social subordinance that is associated and at the same time high levels of overweight. This
with depression or low mood in children depressed IGF- combination of short stature and high fatness is known
1 production (Aßmann and Hermanussen, in press). It is as the nutritional dual-burden. In principle, any group of
possible that IGF-1 is a biomarker which plays an people that has enough energy intake to grow in fatness
important role in the development of social differences should also have enough energy intake to also grow in
in height, body mass and body composition. stature. But, the Maya of Mexico and Central America
It is also likely that the opportunities for physical remain stunted. The Maya are the most numerous of
activity in play, physical education, and paid labor for Native American peoples, with between 7–8 million
the different social classes of these British boys between Maya alive today. In rural areas of Mexico and
the ages of 11 and 16 years had a strong effect. The Guatemala the rates of stunting for the Maya exceed
lower social classes likely gained more lean tissue, 70% of all people.
especially muscle, than did the upper social classes. This For one of our projects we recruited a sample of 57
would raise BMI, but the BMI cannot tell us if this is the urban Maya schoolchildren, aged 7–9 years (31 boys),
case. As a prominent advocate of the use of BMI writes, and 53 of their mothers, mean age 34.44 (sd = 6.3) years.
“But BMI is actually less than ideal for measuring All of the children and their mothers underwent
obesity, as it fails to distinguish between fat mass and anthropometric assessments as well as bioelectrical
muscle mass. When the incidence of obesity first started impedance analysis (BIA). The use of BIA is considered
rising, it is likely that the increase in fat mass was a reliable and accurate method to assess percent body
masked by a corresponding reduction in muscle mass. fatness (%BF) under fieldwork conditions. Multiple
This is particularly true for child obesity, where reduced linear regression was performed to determine whether
physical activity, notably time spent watching the ability of BMI to predict variation in other adiposity
television, is an important risk factor for obesity. So, the indicators is altered by stunting and sitting height ratio
rise in child obesity probably started earlier than 1980, (SHR = sitting height × 100 /total height). The adiposity
though BMI did not reflect it until later” (Cole 1993, p indicators we used were waist circumference (WC), sum
165). If this is true, then BMI is not even a good of the triceps and subscapular skinfolds (SSF), upper
screening tool when used for large population surveys as arm muscle area (UAMA), upper arm fat area (UAFA),
it fails to detect changes in fatness until it is ‘too late’ to and arm fat index (AFI).
take preventative action. We found that 18 (31.6%) of the children were
Current research in Mexico by our research team stunted. In all children, BMI significantly predicted
reveals a final ‘bad to horrid’ example of BMI. This measures of abdominal fatness (WC) and total body
example is based on our article, “How Useful Is BMI in adiposity (%BF, SSF) but not peripheral adiposity
Predicting Adiposity Indicators in a Sample of Maya (UAFA, AFI). Stunting status did not modify the power
Children and Women with High Levels of Stunting?” of BMI to predict adiposity indicators. Relative leg
(Wilson et al. 2011). The applicability of BMI to length neither significantly moderated nor mediated the
populations with high levels of stunting has been effect of BMI on adiposity outcomes. These findings
questioned. Stunting refers to short stature for age and suggest that BMI is an appropriate tool to estimate total
when it occurs in groups of people the stunting is usually and central adiposity in this sample of 7–9-year-old
caused by inadequate nutritional balance and/or lack of children, but that BMI fails to predict their fatness when
specific essential nutrients. Stunted people can have low only arm anthropometry is measured. This is an
levels of body fat, but normal amounts of muscle tissue important finding because in practice the most common

8 Bulletin de la Société Suisse d’Anthropologie 18 (2012)

 The Body Mass Index

Fig. 1: Mean height of boys born in March 1958 in England, Scotland, Fig. 2: Mean weight of boys born in England, Scotland, and Wales by
and Wales by social class of their father (or male head of the social class of their father (or male head of the household).
household). Original figures based on data from Lasker and Original figures based on data from Lasker and Mascie-Taylor
Mascie-Taylor (1989, 1996). (1989).

anthropometric measures of nutritional status and health

are height and weight, and then arm circumference and
triceps skinfold. A lack of correspondence between
height and weight, used to calculate BMI, and arm
anthropometry would lead to incorrect assessments and
ineffective interventions to improve health.
In women, BMI significantly predicted abdominal
adiposity (WC) but not peripheral (AFA, AFI) or total
body adiposity (%BF). Stunting independently predicted
a higher %BF, but did not change the association
between BMI and adiposity indicators in any regression
model. Relative leg length was neither significant nor
altered the association between BMI and any adiposity
indicator. BMI appears to be appropriate for use in these Fig. 3: Mean BMI scores of boys born in England, Scotland, and
adult urban Maya women only to predict abdominal Wales by social class of their father (or male head of the
household). Original figures based on data from Lasker and
adiposity. Maternal %BF as measured by bioelectrical
Mascie-Taylor (1989).
impedance (BIA) was not well predicted by the BMI.
The Maya women participating in our study had BMIs
in the range of 25–29.99, which suggests overweight but
not obesity. However, the %BF of these women was
very high with a mean of 42% as measured by BIA. As
such, it is clearly not appropriate to use BMI alone to wealthy, heavily industrialized nations of Western
predict %BF in this sample of adult urban Maya women. Europe and North America. When used for this purpose
If the BMI is used, then it grossly underestimates the the BMI is, generally, a good tool to estimate
levels of obesity in these women and may eliminate overweight. It may be a good tool to estimate fatness in
them from programs to lower body fatness and improve these same groups of people since the advent of the
health. obesity epidemic, that is, since about the year 1980. The
BMI is a bad tool when used to estimate fatness prior to
1980 because BMI cannot distinguish between fat and
Conclusion lean tissue and there was, generally, lower levels of
fatness in the general population before that date. The
The Body Mass Index was developed to estimate the BMI is also a bad tool when used to estimate fatness for
risk for overweight in large samples of people from the individuals in any nation or in any group of people. The

Bulletin der Schweizerischen Gesellschaft für Anthropologie 18 (2012) 9

 B. BOGIN & I. VARELA-SILVA

BMI was never intended to be used for individual Garn SM, Leonard WR, Hawthorne VM 1986. Three
diagnosis. The BMI becomes a horrid tool to estimate limitations of the body mass index. American Journal of
fatness or health risk when used in some groups of Clinical Nutrition 44: 996–997.
people, such as serious athletes, body building Keys A, Fidanza F, Karvonen MJ, Kimura N, Taylor HL 1972.
enthusiasts, people engaged in jobs with strenuous Indices of relative weight and adiposity. Journal of
physical activity, and in groups suffering from the Chronic Diseases 25: 329–343.
nutritional double-burden of short stature with high Khosla T, Lowe CR 1967. Indices of obesity derived from body
body fatness. The Maya people of Merida, Mexico are weight and height. British Journal of preventive and social
medicine 21: 122–128.
just one example of this last problem with the BMI.
Unfortunately, the majority of low income people in the Kumari M, Tabassum F, Clark C, Strachan D, Stansfeld S,
lesser developed nations of the world are suffering from Power C 2008. Social differences in insulin-like growth
factor-1: findings from a British birth cohort. Annals of
or at risk to the nutritional double-burden. This makes
Epidemiology 18: 664–670.
the BMI a very poor instrument for epidemiological
Lasker GW, Mascie-Taylor CGN 1989. Effects of social class
assessment and the apportionment of health intervention
differences and social mobility on growth in height, weight
resources for the most at-risk segments of the world’s and body mass index in a British cohort. Annals of Human
population. Biology 16: 1–8.
Lasker GW, Mascie-Taylor CGN 1996. Influence of social
class on the correlation of stature of adult children with
that of their mothers and fathers. Journal of Biosocial
References Science 28: 117–22.
Aßmann C, Hermanussen M (in press). Modeling Norgan NG, Jones PR 1995. The effect of standardising the
determinants of growth: Evidence for a community-based body mass index for relative sitting height. International
target in height? Pediatric Research. journal of obesity and related metabolic disorders 19:
206–208.
Billewicz WZ, Kemsley WFFF, Thomson AM 1962. Indices of
adiposity. British Journal of preventive and social Quetelet A 1832. Recherches sur le poids de l’homme aux
medicine 16: 183–188. different âges. Nouveaux Memoire de l’Academie Royale
des Sciences et Belles-Lettres de Bruxelles, t. VII.
Bogin B 1999. Patterns of Human Growth. 2nd edition.
Cambridge University Press. Cambridge. Razak F, Anand SS, Shannon H, Vuksan V, Davis B, et al.
2007. Defining obesity cut points in a multiethnic
Bogin B, Beydoun N 2007. The relationship of sitting height
population. Circulation 115: 2111–2118.
ratio to body mass index and fatness in the United States,
1988–1994. In: Singh SP,Gaur R (eds.). Human Body Romero-Corral A, Montori VM, Somers VK, Korinek J,
Composition. Human Ecology Special Issue No. 15. Thomas RJ, Allison TG, Mookadam F, Lopez-Jimenez F
Kamal-Raj. Delhi, 1–8. 2006. Association of bodyweight with total mortality and
with cardiovascular events in coronary artery disease: a
Bogin B, Varela-Silva MI 2010. Leg length, body proportion,
systematic review of cohort studies. The Lancet 368:
and health: a review with a note on beauty. International
666–78.
Journal of Environmental Research and Public Health 7:
1047–1075. Ross WD, Eiben OG 2002. BMI: A critique of its use in human
biology and the health professions. In: Eiben OG, Bodzsár
Bogin B, Varela-Silva MIO 2008. Fatness biases the use of
ÉB (eds.). Children and youth at the beginning of the 21st
estimated leg length as an epidemiological marker for
century. Humanbiologia Budapestinensis 27: 47–56.
adults in the NHANES III sample. International Journal of
Abstract available at http://ludens.elte.hu/~anthrop/
Epidemiology 37: 201–209.
HB27_Ross.pdf
Deurenberg P, Deurenberg-Yap M 2003. Validity of body
Varela-Silva MI, Frisancho AR, Bogin B, Chatkoff D, Smith
composition methods across ethnic population groups.
PK, Dickinson F, Winham D 2007. Behavioral,
Forum of Nutrition 56: 299–301.
environmental, metabolic and intergenerational
Dublin LI, Lotha AJ. Twenty-five years of health progress. components of early life undernutrition leading to later
Metropolitan Life Insurance Co. 1037. New York. obesity in developing nations and in minority groups in the
Eknoyan G 2008. Adolphe Quetelet (1796–1874) – the USA. Collegium Antropologicum 31: 39–46.
average man and indices of obesity. Nephrology Dialysis Wilson HJ, Dickinson F, Griffiths PL, Azcorra H, Bogin B,
Transplantation 23: 47–51. Varela-Silva MI 2011. How useful is BMI in predicting
Florey CV 1970. The use and interpretation of ponderal index adiposity indicators in a sample of Maya children and
and other weight-height ratios in epidemiological studies. women with high levels of stunting? American Journal of
Journal of Chronic Diseases 23: 93–103. Human Biology 23: 780–789.

10 Bulletin de la Société Suisse d’Anthropologie 18 (2012)

 The Body Mass Index

Figures origin
Fig. 1
Original figures based on data from Lasker and Mascie-
Taylor (1989, 1996).

Figs 2–3
Original figures based on data from Lasker and Mascie-
Taylor (1989).

Address:
Barry Bogin
Centre for Global Health and Human Development
School of Sport, Exercise and Health Sciences
Loughborough University
United Kingdom LE11 3TU
Email: [email protected]
Tel.: (0)1509 228819

Bulletin der Schweizerischen Gesellschaft für Anthropologie 18 (2012) 11