In this study, the concordance between the direct and indirect measurements for facial and cranial indices was poor, regardless of age, gender and socioeconomic stratum (Kappa index: 0.189 and 0.388, respectively).
Some craniofacial morphology characteristics are associated with certain malocclusions; therefore, they provide the clinician with valuable information for defining a particular treatment plan. The facial type is an instrumental factor for orthodontic treatment, because it can impact the anchorage system, predicts the growth of maxillo-mandibular structures, muscle strength and stability of treatment . During growth process, cranial and facial development can be influenced by a variety of factors, such as: environmental conditions, socioeconomic stratum, race, ethnicity, breathing pattern and nutritional habits [7, 14, 15]. For example, children from brachycephalic parents show a decreased index when moved to a different country . In addition, in order to establish orthodontic therapy, two basic factors should be considered: 1) assessment of face dimensions: ¿ is the face long or short, leptoprosopic, mesoprosopic or euryprosopic ?, and 2) when performing the intervention, ¿ is a rotational change that may increase or decrease the expression of the dimensions of the face going to be produced?
This study found that the direct classification method for facial index yielded the mesoprosopic and leptoprosopic types as the most predominant, with a percentage of 47.9% each. When compared with other populations, Chileans exhibit a facial index similar to our study ; however, the most predominant facial type among Africans is the leptoprosopic . Regarding the cranial type, the greatest percentage of students presented dolichocephalic (40.9%) and mesocephalic (40.6%), compared with other populations. In Africa, the most prevalent cranial type is the dolichocephalic (66.82%) ,whereas in Southern Iran, the mesocephalic type is the most common (41.98%) , and, in India, the brachycephalic type prevails . Comparing studies in growing children, Indian population presents mesocephalic index (77.92%) in males, and brachycephalic index (80.85%) in females , while Poland children were brachycephalic (81.45%) , as Japanese population . In Iran, 38.6% were euryprosopic and 38% brachycephalic .
In the present study, 21.4% of children were found to be dolichocephalic and leptoprosopic, a result which relates to findings reported in the literature, where face anatomy can be determined by the cranial base acting as a framework [13, 24]. The growth pattern of cranial and facial indices with growth and structural characteristics of the face exhibits some relationship, which is important to know in order to define an interceptive or corrective orthodontic treatment. The dolichocephalic shape is associated with a leptoprosopic facial type. In contrast, the brachycephalic shape corresponds with an euryprosopic facial type . In dolichocephalic individuals, the brain is relatively narrow and elongated sagittally; this establishes a flatter cranial base, i.e., the angle between the middle and anterior cranial base is wider, which has the following basic implications for the face pattern: 1) the entire naso-maxillary complex is moved forward relative to the jaw due to the rotation of the skull base to the front, and the anterior and middle segments of the cranial base are elongated sagittally; 2) the entire nasomaxillary complex is lower in relation to the mandibular condyle; this causes a downward and backward rotation of the mandible. These people tend to have a retrognathic profile [26, 27].
The euryprosopic facial type was the least frequent in this population. Individuals with this type of face generally have strong muscles and morphological features, such as larger transverse size and parallelism between the occlusal and mandibular planes, smaller gonial angle and decreased lower anterior facial height . Those individuals with a brachycephalic type are usually Class III individuals due to a more posterior position of the maxilla, and have a more anterior location of the mandible . However, it is necessary to note that some individuals may present compensations counteracting some malocclusion trends associated with different skeletal types [27–29]. Confirmation of those relationships could be important to determine in a future study.
This illustrates the need to obtain enough data of all individual traits in order to define an adequate therapy plan aiming to reach aesthetic, functional and stability dental-related goals .
In this study, a poor level of agreement was found between the direct (anthropometer) and indirect (visual) measurement for both facial and cranial indices. One of the reasons that may explain this finding may be a variation in the location of some landmarks, since it requires palpation representing an underlying skeletal structure; and for the visual assessment, this is made from a superficial soft- tissue, which probably leads to differences in the distances between these points . This may be due to the shape of soft- tissue, which correlates approximately 50% to the form of hard tissue , leading to diagnostic mistakes. Another possible explanation is that the percentages used for craniofacial classification were estimated in populations from Europe [12, 32], where the Caucasian race is most predominant and exhibits morphological features different from ours (in Colombia, and mainly in Antioquia, where a mixed race prevails). In 2007,  Farkas et al. reported significant differences in anthropometric measurements of the craniofacial complex, particularly for the orbit and the nose areas, among white individuals from North America and Afro-Americans between 18–25 -year-old. The authors conclude that there must be separate standards for both groups in order to determine intervention guidelines for the head and face. In clinical practice, decisions hinge on the international anthropometric study of the facial morphology conducted among healthy individuals from Europe (Caucasian), Middle East, Asia and Africa. This study determined face anthropometric measurements (18- 30- year- old). However, it is necessary to establish facial and cranial measurements in our population, in order to be able to define morphological alterations, and to adequately intervene for surgical correction . Additionally, Farkas et al.  compared the differences between direct (anthropometrics) and indirect (cephalometries) measurement of dry skulls, and found that the measurements obtained from X-rays were significantly smaller than those obtained from the skull surface. Weinberg et al.  compared accuracy between direct anthropometrics and 3D images and did not find significant differences. One of the advantages of the 3D scanning method is that angles, surface areas, volumes and linear distances can be quantified. One of the drawbacks of this method, which is not commonly mentioned, is its high cost.
One of the benefits of direct assessment is the fact that accurate measurements can be obtained without the need to expose patients to ionizing radiation. Also, cost overruns can be avoided, since the only tool needed is an anthropometer. Another advantage observed in the present study was the short time it took to perform direct measurement (about 2 minutes or less to conduct the four measurements necessary for the two indices: facial and cranial), as opposed to other studies which assert that conducting a greater number of measurements in the course of the same exam could become arduous and difficult .
An additional strength of this study was the population enrolled: 8 -15- year- old children. In this age range, the most interceptive and corrective treatments are performed , since most changes in the craniofacial complex occur during this period. Such changes are first completed in the skull, followed by face width, face depth, and finally, the face length. Due to this facts, they are most likely to be modified . This confirms Farkas assertion, highlighting the importance of developing more detailed anthropometric databases for each ethnical group, due to differences in individuals morphology , and for this current case, not only because of racial differences, but also the age group; for example, Russians kids are brachycephalic, while Americans are mesocephalic .
According to the literature reviewed, both in Spanish and in English, until it is known, this is the first study that compares a direct measurement method (anthropometer on individuals as opposed to dry skulls) with an indirect method (clinical assessment). Previous studies focused primarily on assessing accuracy between direct and indirect measurement techniques (one of them on dry skulls), which compared photographs, cephalometries or 3D images [31, 34, 35].
One drawback of this study is that due to its cross-section nature, associations between craniofacial changes taking place during the growth process and the indices could not be established. To do so, a cohort study would be required, in order to determine whether this classification of individuals is modified as their development stage is completed.