Annals of Human Biology, May–June 2005; 32(3): 390–396

SHORT REPORT

A cross-sectional study of human craniofacial growth

Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

MARINA L. SARDI & FERNANDO V. RAMÍREZ ROZZI

Dynamique de l’Evolution Humaine, Centre National de la Recherche Scientifique, Paris, France

(Received 7 July 2004; revised 22 November 2004; accepted 29 November 2004)

Abstract
It is generally accepted that different cranial regions do not follow the same growth pattern. In this
study, size changes of the functional cranial components (FCCs) in 228 human skulls of age at
death between 0 and 20 years were evaluated. The skull is considered as divided into anteroneural,
For personal use only.

midneural, posteroneural, otic, optic, respiratory, masticatory and alveolar FCCs. Age-related changes
of FCCs were assessed by fitting curves with the smoothing spline method, and quantifying the
proportional increments at different stages. All FCCs show a high growth rate in the first 3–5 years
of life. Two groups of growth trajectories can be distinguished. The anteroneural, midneural,
posteroneural and optic FCCs are more advanced at all stages; they show a high growth rate before
3–5-years-old and a low rate later. This difference is less pronounced in the group comprising
the respiratory, masticatory and otic FCCs. The alveolar FCC shows an independent pattern. The
similarities among FCCs of the two groups are best explained by their common embryological
origin. In contrast, the participation in a common function cannot be associated with the co-ordinated
variation, given that the masticatory and alveolar FCCs show independent trajectories.

Keywords: Functional cranial components, growth pattern, smoothing spline

Introduction
Different growth patterns for neurocranium and face have been described (Corner
& Richtsmeier 1992, Enlow & Hans 1996, Humphrey 1998), the neural system being the
first to attain adult size, followed by the upper face, the mid-face and the mandible. This
differential growth results from differences in rate and/or duration of growth. Although it
is accepted that, at the individual level, the co-variation among traits is due to functional/
developmental constraints (Cheverud 1996), the associations of cranial regions across
ontogeny are quite unknown.
The Functional Matrix Hypothesis (FMH) states that cranial shape is not the result of
its own genetic control, instead it is epigenetically modified by the tissues and cavities

Correspondence: Marina L. Sardi, UPR 2147, Dynamique de l’Evolution Humaine, Centre National de la Recherche Scientifique,
44, rue de l’Amiral Mouchez, 75014, Paris, France. E-mail: msardi@fcnym.unlp.edu.ar
ISSN 0301–4460 print/ISSN 1464–5033 online/00/000390–07 # 2005 Taylor & Francis Group Ltd
DOI: 10.1080/03014460400027441
 Cross-sectional study of craniofacial growth 391

that perform a function (Moss 1973, 1997). The FMH considers the skull as consisting of
functional cranial components (FCCs). Each FCC includes a functional matrix (soft tissues
and cavities) and a skeletal unit (hard tissues that give mechanical support and protection to
the functional matrix) (Moss 1973).
The aims of the present work are to establish the growth trajectories of FCCs in
individuals of ages between 0- and 20-years-old and to estimate the proportion of growth
attained in each FCC at different stages.
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

Materials and methods

A sample of 228 human skulls of known age at death from 0 to 20 years was measured.
Another 121 human skulls from 21- to 39-years-old with all sutures closed and the complete
permanent dentition were added as adult reference. Sex was known for most of individuals
and they were balanced across ages. Those of unknown sex were mostly concentrated
between 0 and 5 ages. The skulls are housed at the Museu Antropologico of Coimbra
(Portugal) and the Musée de l’Homme of Paris (France).
Landmarks and measurements were performed following the method, based on the
FMH, previously applied (Pucciarelli et al. 1990, Sardi et al., 2004). The neurocranium
was divided into four FCCs: anteroneural, midneural, posteroneural, and otic. The distinc-
tion of anteroneural, midneural and posteroneural FCCs does not imply that their matrices
have different functions since cognitive and sensory activities are performed by different and
For personal use only.

broadly distributed parts of the brain and cerebellum (MacLeod et al. 2003). The distinc-
tion was mainly made to gain more insight on the dynamics of neurocranial variation.
Four FCCs comprise the face: optic, respiratory, masticatory, and alveolar.
The length, breadth and height were measured on each FCC with the Microscribe
digitizer. Volumetric indices (VI), representing the geometric mean of the three dimensions
were constructed to estimate size variation, as follows
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VI ¼ 3 ðlength
breadth
heightÞ ð1Þ
Analyses were carried out pooling all individuals, regardless of sex. This enabled us to
include those of unknown sex which belong to the most critical period of growth. Growth
trajectories of FCCs were fitted with the non-parametric smoothing spline. Non-parametric
methods are preferable to describe growth because curves are estimated without the
constraint to fit a particular shape. The spline method consists of dividing the X range in
segments and then adjusting Y values with a parametric regression within each segment.
Thus, fittings of Y in a segment are independent of fittings in adjacent segments. The
smoothing spline requires the definition of the smoothing parameter l which establishes the
trade-off between the bias and the variance. A small l produces a low smooth and a high
variance and a high l produces a good smooth with low variance. Some l were explored but
l ¼ 10 and 100 were chosen by visual inspection.
The quantity of growth of FCCs was measured at ages 0, 7, 14 and 20. This division
enables modifications to be seen across long periods and it is supported by some develop-
mental changes. The ages 0 and 20 represent the beginning and the end of the growth
period; at 7-years-old some behavioural changes, such as the acquisition of an adult diet,
have occurred (Bogin 1997) and the first molar emerged (Smith et al. 1994); at 14-years-
old the pubertal growth spurt and the emergence of second molar occurred in both sexes
(Smith et al. 1994). Considering the averages of adult volumetric indices as a reference,
growth rate was evaluated as the percentage of adult size attained at ages 0, 7, 14 and
 392 M. L. Sardi & F. V. Ramı´rez Rozzi

20 in each FCC. The quantity of changes between two ages was also calculated as a relative
difference between the averages of volumetric indices at the beginning and at the end of each
stage, as follows:
Average at the end of the stage  Average at the beginning of the stage

100 ð2Þ
Average at the beginning of the stage

Results
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

As seen in Figure 1 both ls, 10 and 100, produce very close adjustments. Differences are
mainly located along the first 5 years of life. A l ¼ 10 produces an inflection point around
2–3 years while a l ¼ 100 produces it around 4–5 years in all trajectories. According to
r2 values (Figure 1) both ls explain a high quantity of variation.
Some characteristics of craniofacial growth are evidenced. Two suites of FCCs can be
distinguished by their similarity in trajectories. The first suite comprises anteroneural,
midneural, posteroneural and optic FCCs. Growth trajectories are characterized by a
rapid increment up to 3 or 5 years depending on l, arriving near the adult size. An acute
inflection point is observed at this age, followed by a low growth rate. The second suite
comprises the respiratory, masticatory and otic FCCs. They present a rapid increment
but less pronounced than in the first suite. The inflection point is less acute and after
that, the growth rate is less important. The alveolar FCC presents a different trajectory
For personal use only.

with two periods of greater growth rate, about 0–4 years and after 14 years.
Table I shows the proportion of adult size at ages 0, 7, 14 and 20. Most of the FCCs
present about 50–60% of adult size at birth, the midneural FCC being the most advanced
and the masticatory FCC the most retarded. At ages 7 and 14, the neural FCCs are more
advanced than the facial ones, the masticatory and alveolar being the most retarded.
Concerning the percentage of increment between two stages, at age 7 all the FCCs have
grown about 50% and 68% of the birth size. The masticatory FCC duplicated its size
(116%), indicating that it is the most dynamic. Between ages 7 and 14 the otic, respiratory,
masticatory, and alveolar FCCs are the most dynamic, showing increments about 15%
of the size at age 7. Between ages 14 and 20, the alveolar and masticatory FCCs underwent
the greatest increments. The relative changes of FCCs provide insight into differential
growth. The masticatory FCC shows the greatest growth rate across postnatal life.

Discussion and conclusions

Functional and developmental constraints are considered as a principal source of co-ordinate
variation of skull regions (Cheverud 1996). The similarities in growth trajectories among
FCCs of each suite as well as their advancement regarding adult size can be explained by
their embryological origins. The first suite of FCCs, which contains the brain, cerebellum
and the ocular globe (even when this one is localized in the face), has a common origin in
the neural tube. Functional matrices of the second suite (the cavity for respiration and
smell, muscles for mastication, and structures for hearing and equilibrium) are more hetero-
geneous; however, they present a common embryological origin in the branchial arches.
Regarding functional constraints, it is worth noting that masticatory and alveolar
FCCs, which both intervene in mastication and food treatment, present different growth
trajectories (Figure 1). Function influences the facial development, since the great
increments in the masticatory FCC are undoubtedly associated with the performance of
 Cross-sectional study of craniofacial growth 393
40
100

90 35
Anteroneural
80

Optic
30
70

60 25
r2 λ 100 = 0.927 r2 λ100 = 0.924
50 r 2 λ10 = 0.937 r2 λ10 = 0.933
20
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

0 5 10 15 20 0 5 10 15 20
Age Age

120 40

110 35

Respiratory
Midneural

100 30

90 25

r2 λ100 = 0.907 r2 λ100 = 0.933

80 r2 λ10 = 0.921 20 r 2 λ10 = 0.938

0 5 10 15 20 0 5 10 15 20
For personal use only.

Age Age

50
80

40
Posteroneural

70
Masticatory

60 30

50 r2 λ100 = 0.872 20 r2 λ100 = 0.934

r 2 λ10 = 0.893 r 2 λ10 = 0.938

0 5 10 15 20 0 5 10 15 20
Age Age

16 30

14
25
Alveolar
Optic

12
20
10

r2 λ100 = 0.824 15 r2 λ100 = 0.716

8 r 2 λ10 = 0.828 r 2 λ10 = 0.724

0 5 10 15 20 0 5 10 15 20
Age Age

Figure 1. Smoothing spline of volumetric indices of FCCs vs. chronological age. Solid lines:
smoothness with l ¼ 100; dotted lines: smoothness with l ¼ 10. r2 values for each l measure the
proportion of variation around the lines. The lines perpendicular to the y axes represent the averages
of the adult samples.
 394 M. L. Sardi & F. V. Ramı´rez Rozzi

Table I. Increments of FCCs in different stages expressed as the percentage of adult size, and as relative increments
between two ages according to equation (2).

% of adult size at ages % of increment between ages

FCC 0 7 14 20 0–7 7–14 14–20

Anteroneural 53.6 90.0 96.3 99.3 67.1 7.0 3.1

Midneural 63.4 96.0 96.3 100.0 50.7 0.3 4.6
Posteroneural 58.9 91.8 96.0 98.0 54.6 4.5 2.0
Otic 50.1 81.5 96.0 98.0 63.4 17.9 1.9
Optic 60.5 91.7 95.7 100.0 50.4 4.4 5.9
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

Respiratory 50.7 83.5 94.0 98.4 64.6 12.4 4.8

Masticatory 36.7 79.8 91.3 100.0 116.0 14.6 10.5
Alveolar 42.5 70.0 82.1 99.0 63.4 16.9 20.7

mastication (Herring 1993) but the participation in a common function with the alveolar
FCC appears as a negligible factor in producing a pattern of co-ordinated variation. The
independent growth trajectories can be due to differences in the functional matrices. The
masticatory matrix comprises muscles whereas the alveolar matrix comprises teeth. The pro-
nounced increments in the masticatory FCC up to age 7 (Table I) enable the performance
of mastication after weaning and before the acquisition of permanent dentition. Digestion
of solid food is necessary to satisfy the energy requirements of the brain growth (up to
5 years) and to ensure the body growth until the acquisition of an adult diet after the emer-
For personal use only.

gence of the first permanent molar. The masticatory increments are probably not limited
to digestion, given that mechanical forces increase the blood circulation to the brain
and also influence the circulation of hormones with effects on cognitive competencies
(Nakata 1998).
In conclusion, although a particular growth trajectory is evidenced for the alveolar FCC,
growth trajectories of different regions of the skull can be grouped into two suites according
to the growth rate and according to the changes of growth rate (inflection point). The
similarities among some trajectories indicate that FCCs grow independently of their
localization (neurocranium or face) and the co-ordinate variation seems to be associated
with developmental factors.

Acknowledgements
We want to thank to one of the referees who made useful corrections to the manuscript, and
also to Phillipe Mennecier, Eugenia Cunha and Sofia Wasterlain for access to collections
under their care. This work was made possible by a Fondation Fyssen Grant given to M.L.S.

References
Bogin B. 1997. Evolutionary hypotheses for human childhood. Yearbook Phys Anthropol 40:63–89.
Cheverud JM. 1996. Developmental integration and the evolution of pleiotropy. Am Zool 36:44–50.
Corner BD, Richtsmeier JT. 1992. Cranial growth in the Squirrel monkey (Saimiri sciureus): A quantitative analysis
using three dimensional coordinate data. Am J Phys Anthropol 87:67–81.
Enlow DH, Hans MG. 1996. Crecimiento facial. Mexico DF: McGraw-Hill Interamericana.
Herring SW. 1993. Formation of the vertebrate face: Epigenetic and functional influences. Am Zool 33:472–483.
Humphrey LT. 1998. Growth patterns in the modern human skeleton. Am J Phys Anthropol 105:57–72.
 Cross-sectional study of craniofacial growth 395

MacLeod CE, Zilles K, Schleicher A, Rilling JK, Gibson KR. 2003. Expansion of the neocerebellum in
Hominoidea. J Hum Evol 44:401–429.
Moss ML. 1973. A functional cranial analysis of primate craniofacial growth. Symposia of the IVth International
Congress of Primatology 3:191–208.
Moss ML. 1997. The functional matrix hypothesis revisited. 4. The epigenetic antithesis and the resolving
synthesis. Am J Orthod Dentofacial Orthop 112:410–417.
Nakata M. 1998. Masticatory function and its effects on general health. Int Dent J 48:540–548.
Pucciarelli HM, Dressino V, Niveiro MH. 1990. Changes in skull components of the Skirrel monkey evoked by
growth and nutrition: An experimental study. Am J Phys Anthropol 81:535–543.
Sardi ML, Ramı́rez Rozzi F, Pucciarelli HM. 2004. The Neolithic transition in Europe and North Africa.
The functional craniology contribution. Antropol Anz 62:129–145.
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

Smith BH, Crummett TL, Brandt KL. 1994. Ages of eruption of primate teeth: A compendium for aging
individuals and comparing life histories. Yearbook Phys Anthropol 37:177–231.

Résumé. On s’accorde généralement à admettre que différentes régions crâniennes ne

suivent pas le même schéma de croissance. Cette étude évalue les changements de taille des
composants crâniens fonctionnels (CCF) de 228 crânes dont l’âge au décès varie de 0 à 20
ans. On considère que le crâne est divisé en CCFs anténeuronal, medioneural,
postéroneural, otique, optique, respiratoire, masticatoire et alvéolaire. Les changements
des CCF relatifs à l’âge sont exprimés par des courbes ajustées et lissées, les accroissements
proportionnels à des étapes différentes étant quantifiés. Tous les CCF présentent un rythme
de croissance élevé au cours des 3 à 5 premières années. Deux groupes de trajectoires de
For personal use only.

croissance peuvent être distingués. Les CCF anténeuronal, medioneural, postéroneural et

optique sont les plus avancés à tous les âges ; ils montrent une vitesse de croissance élevée
avant 3-5 ans, puis une vitesse lente par la suite. Cette différence est moins prononcée dans
le groupe qui comprend les CCF respiratoire, masticatoire et otique, alors que le CCF
alvéolaire présente une modalité indépendante. Les similarités des deux groupes de CCF
s’expliquent par leur origine embryologique commune. Par contraste, étant donné que les
CCF masticatoire et alvéolaire présentent des trajectoires indépendantes, la participation à
une fonction commune ne peut pas être associée à une variation coordonnée.

Zusammenfassung. Es ist allgemein akzeptiert, dass verschiedene Regionen des Schädels

nicht denselben Wachstumsmustern folgen. In dieser Studie wurden Größenänderungen
der funktionellen kranialen Komponenten (FCC) an 228 menschlichen Schädeln
untersucht, mit einem Alter zum Zeitpunkt des Todes von 0 bis 20 Jahren. Der Schädel
wird dabei als in funktionelle Komponenten unterteilt angesehen: anteroneural (vor dem
Hirn liegend, Gesichtsschädel), midneural (Bereich der Schädelbasis) und posteroneural
(hinter der Schädelbasis liegend, Hirnschädel); ferner die Ohrregion (otic), die
Orbitalregion (optic), die Region des Nasenskelettes (respiratory), der Kauapparat
(masticatory), und das Kieferskelett (alveolar FCC). Altersbedingte Veränderungen der
FCC wurden durch Kurvenanpassung mittels glättender spline Funktionen bestimmt und
so die proportionalen Zuwächse zu jedem Alter quantifiziert. Alle FCC zeigen in den ersten
3-5 Lebensjahren hohe Wachstumsraten. Dabei können 2 Gruppen von
Wachstumsabläufen unterschieden werden. Die anteroneuralen, midneuralen und poster-
oneuralen FCC, sowie die Orbitalregion sind in jedem Alter weiter entwickelt; sie zeigen
eine hohe Wachstumsrate vor dem Alter von 3-5 Jahren und eine niedrige Rate danach.
Dieser Unterschied ist weniger ausgeprägt in der Gruppe von FCC, die aus dem
Nasenskelett, dem Kauapparat und der Ohrregion bestehen. Das Kieferskelett zeigt ein
unabhängiges Muster. Die übereinstimmungen der FCC beider Gruppen können am
 396 M. L. Sardi & F. V. Ramı´rez Rozzi

besten mit ihrer gemeinsamen embryonalen Herkunft erklärt werden. Dagegen kann die
Beteiligung an einer gemeinsamen Funktion nicht über diese geordnete Variation erklärt
werden, denn Kauapparat und Kieferskelett zeigen voneinander unabhängige
Wachstumsabläufe.

Resumen. En general, se acepta que las diferentes regiones craneales no siguen el mismo
patrón de crecimiento. En este estudio, se evaluaron los cambios en tamaño de los
componentes craneales funcionales (CCFs) de 228 cráneos humanos, con edades de muerte
entre 0 y 20 años de edad. El cráneo se consideró dividido en CCFs antero-neural, medio-
Ann Hum Biol Downloaded from informahealthcare.com by Francis A Countway Library of Medicine on 08/23/13

neural, póstero-neural, ótico, óptico, respiratorio, masticatorio y alveolar. Los cambios en

los CCFs relacionados con la edad se estimaron mediante el ajuste de curvas con el método
de esplines suavizados, y cuantificando los incrementos proporcionales en los diferentes
estadios. Todos los CCFs mostraron una elevada tasa de crecimiento durante los primeros
3-5 años de vida. Pueden distinguirse dos grupos de trayectorias de crecimiento. Los CCFs
antero-neural, medio-neural, póstero-neural y óptico están más avanzados en todos los
estadios; muestran una elevada tasa de crecimiento antes de los 3-5 años de edad y una tasa
baja después. Esta diferencia es menos pronunciada en el grupo que incluye los CCFs
respiratorio, masticatorio y ótico. El CCF alveolar muestra un patrón independiente. Las
semejanzas entre los CCFs de los dos grupos se explican mejor por su origen embriológico
común. Por el contrario, la participación en una función común no puede asociarse con la
variación coordinada, dado que los CCFs masticatorio y alveolar muestran trayectorias
For personal use only.

independientes.