Keywords
anatomy and anthropology - anthropometry - body height - forensic anthropology - jaw
- legal medicine
Introduction
An individual's sex, age and height are of great importance in legal medical practice
for identifying corpses, and especially so when bodies are reduced to their skeletons.[1] In these situations, the absence of biological characteristics that would enable
identification of individuals in scenarios of aviation accidents, automobile accidents,
homicides, fires, or natural disasters still constitutes a problem, even in light
of the various technical possibilities that exist within forensic anthropology. Height
is one of the main criteria used in identifying human beings. This is based on the
principle that an individual's height has a defined proportional relationship with
the various parts of the body.[2]
[3] However, in some circumstances, it is not always possible to obtain a complete human
skeleton. Thus, it becomes a challenge to estimate an individual's height and identify
this person when only the cephalofacial region of the skeleton is available for an
anthropological legal medical expert examination.[4] Another situation in which height estimates have been shown to be extremely important
are cases of disease or deformity of the spine, which make it difficult to directly
measure the individual's height.[5]
Anatomical and mathematical methods form the two main tools used in forensic anthropology
for estimating height.[6] The anatomical method consists of measuring and adding together all the skeletal
elements that contribute towards an individual's height, with a correction factor
for soft tissues. The mathematical model consists of derivation of formulae through
linear correlations that exist between given parts of the body and height.[7]
[8] For this calculation, some studies[2]
[4]
[9]
[10] have demonstrated that regression analysis is the most reliable method.
Several authors have attempted to correlate height with different parts of the body.[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27] However, few studies on cephalofacial variables have been conducted.[2]
[3]
[28]
[29]
[30]
[31]
[32]
[33]
Differences in cephalofacial morphology are present between different populations
worldwide. These are determined by environmental and ethnic factors and even by the
customs of these populations.[34] Thus, the regression formulae through which it is sought to estimate the height
of different cranial variables are specific for each sample.
Given the scarcity of studies on Brazilian populations and the need to develop such
studies, the present study had the objective of analyzing the relationships that might
exist between measurements of the mandible (length of the mandibular arch and bigonial
width) and height.
Materials and Methods
This was an anthropometric study in which cross-sectional analytical and descriptive
was analyses were conducted. Data were gathered from the subjects after they had been
given explanations regarding the objectives of the study and had signed a free and
informed consent statement. A direct approach was used, at the participants' homes,
in the same way used in surveys conducted by the Brazilian Institute for Geography
and Statistics (IBGE, in the Portuguese acronym). The instruments used in making the
measurements were a pachymeter (Cerscof, Araquari, Santa catarina, Brasil) an inelastic
measuring tape, and a portable stadiometer (ALTUREXATA, Belo Horizonte, Minas Gerais,
Brasil).
This study was conducted among individuals of both genders, aged 18 to 85 years, who
belonged to a quilombo community, Mussuca, in the municipality of Laranjeiras, state
of Sergipe, Brazil. The municipality of Laranjeiras is located in the northeast of
Brazil, in the state of Sergipe, at about 18 Km from the capital, Aracaju. According
to the IBGE, it had an estimated population of 29,700 people in the year of 2017,
distributed in a land area of 162.273 km2 (https://cidades.ibge.gov.br). The sample comprised 300 individuals (150 women and 150 men). The inclusion criteria
were individuals in good health, of either gender, who were ≥ 18 years of age. Individuals
who presented spinal abnormalities (kyphosis, lordosis or scoliosis), lower-limb amputations,
or any macroscopic malformation in the mandibular region were excluded. The statistical
analysis comprised calculation of means, standard deviation (SD), minimum and maximum
values, Pearson correlation coefficient (r), and multiple linear regression analysis.
It was performed using the Statistical Package for the Social Scienes SPSS®, version 19.0, for Windows (SPSS Inc., Chicago, EUA) and Excel 2010 (Microsoft Corp.,
Redmond, WA, USA). The present study was approved by the Research Ethics Committee
of the Federal University of Sergipe, under protocol number 32813314.1.0000.5546.
Anthropometric Measurements:
Anthropometric Measurements:
-
- Bigonial width: this was measured as the linear distance between the two gonia,
which are points at the external angle of the mandible, in posterior view. This was
measured with the aid of a pachymeter (Cerscof) ([Fig.1]).
-
- Length of the mandibular arch: this was measured in anterior view between the mandibular
angles, following the lower edge of the mandible, using an inelastic measuring tape
([Fig 2]).
-
- Height: this was measured using a portable stadiometer (ALTUREXATA) and consisted
of the greatest distance between the plantar region of the feet and the top of the
head (at the coronal vertex).
Fig. 1 Bigonial width.
Fig. 2 Length of the mandibular arch.
Results
The ages among the sample ranged from 18 to 85 years, with a general mean of 38.8
years and a SD of 15.6 years. The mean for males was 39.1 years and for females, 38.4
years, without any statistically significant difference (p = 0.688). The mean height of the men (170.4 cm) was significantly greater (p < 0.001) that that of the women (158.6 cm). The heights of the men ranged from 150.4
to 190.7 cm, while the heights of the women ranged from 142.6 to 170.6 cm. The mean
difference in height was 11.8 ± 0.7, with a 95% confidence interval (CI) of 10.4 to
13.2. The mean values for both anthropometric facial variables were significantly
larger (p < 0.001) among the men. The length of the mandibular arch showed a mean difference
of 1.3 ± 0.12 with 95% CI of 1.04-1.52. The bigonial width of the mandible showed
a mean difference of 0.56 ± 0.07 with 95% CI of 0.42-0.7 ([Table 1]).
Table 1
Characterization of age and anthropometric variables according to gender in a “quilombo”
community
|
Male (n = 150)
|
Female (n = 150)
|
p
|
|
Mean
|
SD
|
Min
|
Max
|
Mean
|
SD
|
Min
|
Max
|
|
Age
|
39.1
|
16.2
|
18
|
85
|
38.4
|
15.1
|
18
|
83
|
0.688
|
|
Height
|
170.4
|
6.7
|
150.4
|
190.7
|
158.6
|
5.9
|
142.6
|
170.6
|
< 0.001
|
|
Length of mandibular arch
|
22.5
|
1.1
|
19.9
|
25.5
|
21.2
|
1.0
|
17.7
|
23.8
|
< 0.001
|
|
Bigonial width of mandible
|
10.5
|
0.7
|
9.0
|
12.2
|
9.9
|
0.5
|
8.6
|
11.3
|
< 0.001
|
Abbreviations: Min, minimum; Max, maximum; SD, standard deviation; p, significance level.
In the present study, the anthropometric measurements on the mandible showed weak
correlations with height, such that the Pearson correlation coefficients (r) were
low. Analysis on these values showed very low positive correlations between height
and the length of the mandibular arch, both among men (r = 0.177; p < 0.05) and among women (r = 0.271; p = 0.01); and between height and the bigonial
width of the mandible among women (r = 0.169; p < 0.05) ([Table 2]).
Table 2
Correlation of height with age and anthropometric variables of the mandible according
to gender
|
Male (n = 150)
|
Female (n = 150)
|
|
r
|
p
|
r
|
p
|
|
Age
|
-0.414
|
< 0.001
|
-0.433
|
< 0.001
|
|
Length of mandibular arch
|
0.177
|
0.030
|
0.271
|
0.001
|
|
Bigonial width of mandible
|
0.144
|
0.079
|
0.169
|
0.038
|
Abbreviations: p, significance level; r, Pearson correlation coefficient.
Multiple regression models were calculated for males and females, with height as the
dependent variable in both cases. For males, age, length of the mandibular arch and
bigonial width of the mandible were used as independent variables. Among these measurements,
the one that presented greatest explanatory power regarding variation of height was
the bigonial width of the mandible (standardized beta = 0.200) ([Table 3]).
Table 3
Multiple linear regression model for males
|
B
|
Beta
|
p
|
|
Constant
|
132.771
|
−
|
< 0.001
|
|
Age
|
-0.209
|
-0.504
|
< 0.001
|
|
Length of mandibular arch
|
1.109
|
0.179
|
0.028
|
|
Bigonial width of mandible
|
1.993
|
0.200
|
0.016
|
Abbreviations: B, non-standardized partial regression coefficient; Beta, standardized
partial regression coefficient; p, significance level.
Among females, age and the length of the mandibular arch were represented as independent
variables. In this multiple linear regression model, the bigonial width of the mandible
was shown not to be statistically significant (p = 0.756). Thus, the length of the mandibular arch emerged as the variable that best
explained variations of height (standardized beta = 0.259) ([Table 4]).
Table 4
Multiple linear regression model for females
|
B
|
Beta
|
p
|
|
Constant
|
133.1
|
−
|
< 0.001
|
|
Age
|
-0.165
|
-0.425
|
< 0.001
|
|
Length of mandibular arch
|
1.456
|
0.259
|
< 0.001
|
|
Bigonial width of mandible
|
0.281
|
0.025
|
0.756
|
Abbreviations: B, non-standardized partial regression coefficient; Beta, standardized
partial regression coefficient; p, significance level.
The regression formulas were calculated separately for each gender, from their independent
variables ([Table 5]). Among the men, age, length of the mandibular arch and bigonial width of the mandible
explained around 27% of the variation of height. Among the women, age and length of
the mandibular arch emerged as the main predictors for height, explaining around 25.4%
of its variation. The standard error of the estimate (SEE) for the regression model
derived was greater for men than for women.
Table 5
Regression equations for height estimates according to gender
|
Sex
|
Equation
|
R2
|
SEE (cm)
|
|
Male
|
Height (cm) = 132.771- 0.209 x age (years) + 1.109 x LMA (cm) + 1.993 x BWM (cm)
|
0.269
|
5.8
|
|
Female
|
Height (cm) = 133.1-0.165 x age (years) + 1.456 x LMA (cm)
|
0.254
|
5.1
|
Abbreviations: BWM, bigonial width of the mandible; LMA, length of the mandibular
arch; R2, coefficient of determination; SEE, standard error of the estimate.
Discussion
So far, anthropometric studies of the Brazilian population with the aim of correlating
the length of the mandibular arch and its bigonial width have been scarce or nonexistent.
Starting from this observation, we sought, with the present study of a population
in a quilombo settlement, to construct a specific equation for each gender that would
be useful for making height estimates from each individual's age and facial variables.
The word “quilombo” in the Yoruba language, means “dwelling”, and Bantu, “gathering
camps,” “union”.[35] For the National Institute of Colonization and Agrarian Reform (INCRA, in the Portuguese
acronym),[36] quilombo communities represent ethnic groups predominantly made up of black population,
self-defined from the relationship with the land, ancestry, territory and own cultural
practices (http://www.incra.gov.br/quilombola).
According to Moura,[37] quilombo is defined as a sociopolitical organization that originated from the fight
and resistance against the slavery system. The quilombos were regarded as a cultural
space of resistance and preservation of the African culture, in which blacks were
able to speak their language and worship their religion; in that sense, the quilombo
represented rebellion against standards imposed by the official society and restoration
of the old values.[38]
The results from this study demonstrated that the length of the mandibular arch presented
the best correlation coefficient with height, both for men (r = 0.177) and for women
(r = 0.271). Previous studies did not mention this facial variable among their results.[3]
[4]
[29]
[30]
[31]
[32]
[33]
[34]
[39]
[40]
[41]
[42] The correlation between the bigonial width of the mandible and height among men
was not shown to be statistically significant (p > 0.05), and this was also observed by Agnihotri et al,[3] and Sahni et al.[4] Among women, the bigonial width presented a weak statistically significant correlation
with height, similar to the values found by Ahmed and Taha,[29] and by Shah et al.[30] A comparison of the different correlation coefficients between bigonial width of
the mandible and height, according to gender, as described by some authors, is presented
in [Table 6].
Table 6
Correlation coefficients (r) between the bigonial width of the mandible and height
in various studies
|
Authors
|
Population
|
Age
(years)
|
Bigonial width of the mandible
|
|
Men
|
Women
|
|
n
|
r
|
p
|
n
|
R
|
p
|
|
Present study
|
Brazilian
|
18–85
|
150
|
0.144
|
0.079
|
150
|
0.169
|
< 0.05
|
|
Ahmed and Taha[29]
|
Sudanese
|
18–25
|
120
|
0.247
|
< 0.01
|
120
|
0.368
|
< 0.001
|
|
Shah et al[30]
|
Indian
|
21–50
|
676
|
0.096
|
< 0.05
|
225
|
0.193
|
< 0.05
|
|
Agnihotri et al[3]
|
Indian-Mauritian
|
20–28
|
75
|
0.022
|
0.853
|
75
|
0.159
|
0.175
|
|
Pelin et al[34]
|
Turkish
|
18–45
|
286
|
0.164
|
< 0.001
|
−
|
−
|
−
|
|
Sahni et al[4]
|
Indian
|
18–70
|
173
|
0.064
|
0.201
|
127
|
0.047
|
0.299
|
|
Krishan[40]
|
Indian
|
18–30
|
996
|
0.462
|
< 0.001
|
−
|
−
|
−
|
|
Krishan and Kumar[31]
|
Indian
|
12–18
|
252
|
0.449
|
< 0.001
|
−
|
−
|
−
|
Abbreviations: n, number of individuals in the sample; p, significance level; R, Pearson correlation coefficient.
Only a few studies have attempted to correlate height with cephalofacial anthropometric
variables in different populations around the world. A study conducted by Krishan
and Kumar[31] on an endogamous group of castes in northern India demonstrated that the 16 cephalofacial
measurements used presented significant correlations with height. The highest correlation
coefficient among these was with the horizontal circumference of the head (r = 0.773),
and the lowest was with the nose size (r = 0.265). A subsequent study conducted by
Krishan[40] on a sample of 996 adults Gujjars in northern India indicate that all the variables
used presented positive correlations with height, with correlation coefficients ranging
from 0.455 to 0.781. In a study on Sudanese Arab students, Ahmed and Taha[29] reported that the highest correlation coefficients between cephalofacial measurements
and height related to the width of the base of the cranium (r = 0.370) and the bizygomatic
width (r = 0.350) among men; and to the bizygomatic width (r = 0.369) and bigonial
width (r = 0.368) among women. Sahni et al[4] found low correlation coefficients between facial variables and height. From measurements
on 124 Japanese cadavers, Chiba and Terazawa[41] found correlation coefficients for various craniofacial parameters ranging from
0.32 to 0.53, but when individuals aged 70 years or over were excluded from the sample,
the correlation coefficients ranged from 0.38 to 0.6. These authors also stated that
height estimates from these variables could be used in forensic practice. In the present
study, the means for all the variables used were shown to be greater among males than
among females. Similar results were obtained by Shah et al[30] and by Patil and Mody[42] in a lateral cephalometric study on adults in central India. Agnihotri et al[3] studied height estimates from cephalofacial variables among 150 young adults aged
20 to 28 years and observed that among the men, the horizontal circumference of the
head (r = 0.494), nose width (r = 0.380) and morphological length of the face (r = 0.328)
stood out as the main factors responsible for height estimates. On the other hand,
among the women, the main predictors were the physiognomic length of the face (r = 0.382),
horizontal circumference of the head (r = 0.375) and bizygomatic width (r = 0.276).
In a Turkish population, Pelin et al[34] sought to evaluate the correlation coefficients between cephalofacial anthropometric
variables and height, according to different types of head (dolichocephalic, mesencephalic,
brachycephalic and hyperbrachycephalic) and different types of face (hypereuryprosopic,
euryprosopic, mesoprosopic, leptoprosopic and hyperleptoprosopic). They concluded
that these variables were not good predictors for estimating height. In a study on
200 cadavers (148 male and 52 female) from a population in Nepal, Shrestha et al[33] found statistically significant correlation coefficients between all the anthropometric
variables and height, for their entire sample. Height in young individuals grows progressively,
reaching the highest values between 21 and 25 years of age. After this age, the values
decrease by an average of 2.5 cm every 25 years. This has been attributed to the wear
of the intervertebral discs, plus a greater postural complacency, to lean, resulting
from the reduction of muscle tone inherent in advancing age.[43] Therefore, it can be observed that the coefficient of Pearson found in our study,
as occurred in most descriptions of the studied literature, is negative when correlated
with age stature, demonstrating objectively that, in fact, there was a decrease in
stature with advancing age of the individual.
This large variability in the results presented, between different studies, in attempts
to find positive and statistically significant correlation coefficients between cephalofacial
anthropometric variables and height may be related to the samples that different researchers
have used (castes versus ethnically mixed populations; and living individuals versus
cadavers) and the methodologies used (facial and/or cranial anthropometry; or imaging
examinations). However, we emphasize that the formulae derived in the present study
are specific for the quilombo population of the Mussuca settlement, in the state of
Sergipe, given that the morphometric diversity of crania seems to be influenced by
environmental, racial and nutritional factors.
Conclusion
All the anthropometric variables of the mandible in the quilombo population studied
presented very low correlation coefficients with height. Therefore, it can be concluded
that height estimates may be less accurate in cases in which only the cephalofacial
dimensions are available for forensic medical anthropological examination.
