Methods of Analysis of the Nasal Profile: A Systematic Review with Meta-analysis

The nose is the most prominent structure of the face, influencing facial appearance and profile. Orthodontists have an awareness of facial structures, including nasal morphology, when diagnosing and treatment planning. Maxillofacial surgeons influence facial profile by bimaxillary surgery, improving facial aesthetics and harmony. The aim of this review was to summarize the available methods of analysing nasal morphology and profile, and to assess their complexity. A literature search was conducted in PubMed, Scopus, Web of Science, and Embase using the following search terms: “nasal profile analysis”, “nasolabial angle”, and “nasal profile cephalometric” in order to select studies providing knowledge on correlations between occlusion and nasal development, differences between skeletal classes, ethnic variability, and differences between the sexes. Studies concerning genetic disorders were excluded. Finally, 17 full-text papers were analysed, which pertained to nasolabial angle, or facial profile including the nose. Data concerning methods, ethnic group, reference landmarks used, and measurements made were extracted and placed in tables. Numerous methods of nasal profile analysis can be found in the literature. These methods describe various numbers of parameters, which have influence on facial aesthetic. Nasal parameters are correlated to skeletal class and nasolabial angle, positions of upper incisors, and maxillary inclination.


Background
The nose is the most prominent element of the face, influencing facial appearance and profile [1][2][3][4][5]. According to the study by Ghorbanyjavadpour and Rakhshan [6], there are some factors associated with the esthetics of the soft-tissue profile, also associated with the nose, such as less prominent noses with higher tips and subnasales anterior to the upper lip. Numerous authors deal with analysis of the facial profile [7][8][9][10][11][12][13][14][15]. The studies take into consideration age, sex, skeletal class, and ethnic group. Maxillofacial surgeons influence facial profile by bimaxillary surgery, improving facial aesthetics and harmony [16], as well as nasal projection and nasolabial angle (NLA) [17]. Nasal growth since early childhood as well as nasal shape and profile has been subjected to various analyses by numerous authors [1-3, 7, 9, 14, 18-22]. Orthodontists have an awareness of facial structures, including nasal morphology, when diagnosing and treatment planning in order to achieve good results after treatment cessation [11]. The aim of this review was to summarize the available methods of analysing nasal morphology and profile and to assess their complexity.

Material and Methods
2.1. Search Strategy. The search and the entire review were performed according to the PRISMA statement [23] and following the guidelines from the Cochrane Handbook for Systematic Reviews of Interventions [24]. All searching was performed using a combination of different MeSH terms and free-text terms. After all, the final search strategy was determined by several presearches. The literature search was conducted in following databases: PubMed, Scopus, Web of Science, and Embase using the following search terms: "nasal profile analysis" OR "nasolabial angle" OR "nasal profile cephalometry" on 6 th December 2020. The papers initially selected were subjected to detailed analysis, regarding the methods of analysis as well as knowledge on nasal morphology and development.
Then, the following exclusion criteria were employed for this review: (1) case reports; (2) reviews; (3) abstract and author debates or editorials; (4) lack of effective statistical analysis; (5) studies concerning congenital deformities; (6) studies evaluating theoretical algorithms, classification systems, or descriptions of protocols. No limitation referring to the year of publication of the studies was imposed.
All papers found were analysed in order to select studies providing knowledge on correlations between occlusion and nasal development, differences between skeletal classes, ethnic variability, and differences between the sexes.

Data Extraction.
Titles and abstracts were selected independently by two authors (MJ and AJ), following the inclusion criteria. The full text of each identified primarily included article was then analysed to find out whether it was appropriate for inclusion. Disagreements were resolved through discussion with the team supervisor (JJO). Authorship, year of publication, data concerning methods, ethnic group, reference landmarks used, and measurements taken were independently extracted by two authors (AJ and MJ) and examined by the third author (JJO).

Risk of Bias.
According to the PRISMA statements, the evaluation of methodological quality gives an indication of the strength of evidence provided by the study because methodological flaws can result in bias [23]. For studies based on the observation of structures found in radiological examinations, a specific scale for Clinical Studies of Radiologic Examinations should be applied. For this reason, it was decided to use the Arrive´Scale [25]. It consists of 15 components, i.e., study design, study purpose, reference standard, inclusion criteria, indeterminate results, exclusion criteria, spectrum of patients, analysis method, analysis criteria, avoided work-up bias, avoided diagnostic-review bias, avoided testreview bias, intraobserver reliability, interobserver reliability, and statistical analysis, that accurately assess the bias risk, and due to their complexity, they provide detailed analysis of the results. One point is given for the compliance of the test characteristics with the required characteristics listed in the scale. In the event of a defect in the methodology, the research receives 0 points. The more points the research received, the better the methodology it has.
2.5. Meta-analysis. Meta-analysis was performed using random-effects model via metafor and compute.es R packages [26] with Standardized Mean Differences (SMD) and 95% confidence intervals (95% CI) being calculated as effect estimates. Heterogeneity was assessed quantitatively using I 2 -statistics and Cochran's Q. [27]. The meta-analysis included studies that examined the values of nasiolabial angle separately for women and men and provided SD values for both groups.

Results
The search strategy identified 3874 potential articles: 3381 from PubMed, 241 from Scopus, 177 from Web of Science, and 75 from Embase. After duplicates had been removed, 3534 articles were screened. After that, 3493 papers were excluded because they did not correspond with the topic of this review. Of the remaining 41 papers, 24 were excluded because they were not relevant to the eligibility criteria. Finally, 17 full-text papers were included into qualitative analysis ( Figure 1 PRISMA 2009 Flow Diagram). All of included studies pertained to nasolabial angle, or facial profile, including the nose.
The study material of the studies included is presented in Table 1. Table 2 presents nasal and cephalometric landmarks from the literature. Angular and linear variables from the studies included have been described in Table 3. 3.1. Risk of Bias. The Arrive Scale was chosen in order to unify quality assessment of all studies included in this systematic review. If a decision was made to choose less specific, more popular scales, such as Newcastle-Ottawa scale or Jadad scale, it would cause chaos due to overdivision through various types of research, so the results of the risk of bias assessment would not be transparent. (Table 4).

Meta-analysis.
Many of the studies included in the review leave the question open as to whether gender influences the nasiolabial angle. It was concluded that it is worth performing metanalysis in order to unify the results included in the review of studies and draw a common, consistent conclusion. There were 8 included studies in metanalysis. The values and SD of NLA that were reported are presented in Table 5.
The results by Hwang et al. [28], Paradowska-Stolarz and Kawala [29], and Kumar et al. [30] are presented separately (in 2 separate groups) because of the significant factors differing study groups. The results are shown in Figure 2. SMD should be treated as measure of gender influence on the value of nasiolabial angle. Positive value of SMD indicates greater angle in male patients, negative-in female patients.
Forest plot of 11 studies on gender influence on the value of nasiolabial angle has been presented in Figure 2. Positive value of SMD indicates a higher angle in male patients, negative-in female patients. Gender has an insignificant (p = 0:671) negative effect size. Study results are consistent-heterogeneity is insignificant (p = 0:228); only 18.5% of the variability come from heterogeneity. Funnel plot ( Figure 3) does not reveal publication bias.

Discussion
Part of the methodology that is missing in many of the studies included in the review is patient selection and comparability. The number of women and the number of men were in some studies unequal [29,[31][32][33] This deficiency was often caused by the randomization procedure (selection of cephalograms); sometimes, it was a simple negligence of researchers. An important thing, which was also missing in many papers, was the analysis of cephalograms by more than one observer [28,29,31,32,[34][35][36][37]. Even if intraobserver reliability is ensured by performing more than one cephalometric analysis of each radiograph, it is important to have another researcher or even computer AI as verification of identification of cephalometric landmarks. This is noticeable that more recent research places more emphasis on this aspect of the methodology. Due to the common standardization of cephalometric analysis and the additional description of the position of the points in every paper, there were no objections in such parts of evaluation as reference standard, indeterminate results, or in the method and criteria of the analysis. These attributes result from the subject of research (the principles of cephalometric analysis are well known and well-established), not from the way it was conducted.
The present study is a summary of the available methods of analyzing nasal profile on lateral cephalograms. All landmarks and measurements have been tabularized. Moreover, scientific findings on correlations between nasal and craniofacial morphology as well as age, gender, and ethnicity have been summarized. The results of this systematic review may aid treatment planning, when facial and esthetics are the primary goal.

BioMed Research International
Gulsen et al. [38] defined numerous nasal parameters describing shape, size, the presence of dorsal curvatures, and nasal length or depth. The same method of nasal analysis was later used by Arshad et al. [31] on cephalometric radiographs of 119 subjects aged 18-40, with no congenital deformities and no history of orthodontic treatment, in order to assess nasal profile in the sagittal and vertical planes and analyse sexual dimorphism. The variables measured on cephalometric radiographs are presented in Table 3. Nehra and Sharma [34] analysed correlation between vertical skeletal pattern of the maxilla and nasal morphology on 190 pretreatment lateral cephalometric radiographs of patients aged 18-27 years. The cephalometric variables used can be found in Table 3. They found a correlation between nasal parameters (length, depth, tip angle, and nasolabial angle), maxillary and mandibular inclinations, and anterior and posterior facial heights. They found significant correlations between nasal values (length and nose tip angle) and position of the maxilla.
Changes in nasal growth, size, and morphology referring to the vertical pterygomaxillary plane (PMV) were described by Meng et al. [39] based on 305 cephalometric radiographs of 23 females and 17 males aged 7-18 years. The reference landmarks, lines, and angles are presented in Tables 2 and   Table 2: Nasal and cephalometric landmarks in the literature (included in the paper nos. [28][29][30]32).   Moreover, girls had 90% of their final nasal height already at the age of 7, boys after the age of 17. Nasal depth was 70% of its final measure in girls at the age of 7, in boys at the age of 11. Nasal measurements were always lower in girls than in boys. A single study by Buschang et al. [40] was found describing a growth analysis of the upper and lower parts of the nasal dorsum in children and adolescents. The nasal dorsum appeared to grow on average by 10°between 6 and 14 years of age. Its development is correlated with nasal tip growth. In subjects with a horizontal skeletal growth pattern, the dorsum moves upwards and forwards. In the case of vertical growth, rotations occur directed downwards and backwards.

Landmarks
The correlation between skeletal patterns and nasal shape was analysed by Robinson et al. [37] in 123 women, based on two angles and three linear measurements (Table 3). A straight nasal dorsum was more prevalent in skeletal class I, convex (nasal hump)-in class II, whereas concave-in class III. Nasal length and depth were strongly correlated with age.
Skinazi et al. [35] focused on analyzing nasal surface area in a French population. They drew Rickett's Esthetic Line (from nasal tip to the most prominent chin point) and Juanita line (drawn from Sn (the depth of the nasolabial sulcus) to the deepest point of the labiomental sulcus) on facial profile and calculated surface areas of the nose, lips, and chin. Mean nasal surface area was 246:14 ± 65:51 mm 2 in women and 235:4 ± 59:16 mm 2 in men.
Eight papers were found pertaining to NLA. The authors of the studies cited [28,[31][32][33] analysed variability in nasal morphology. The data extracted are presented in Table 5.
Both Arshad et al. [31] and Gulsen et al. [38] found a significant correlation between nasomental angle (NMA) and convexity of facial soft tissue (SFC) and skeletal class. NMA values are higher in skeletal class III, lower in class II [31,38]. A higher SFC angle is found in class II, a lower in class III [31,38]. Moreover, Gulsen et al. [38] found correlation between NLA and skeletal class. Arshad et al. [31] indicated that concavity of the lower part of the nasal dorsum (Dconv) is strongly dependent on skeletal class. Numerous authors proved significant sexual dimorphism concerning individual nasal variables. The results indicate that both nasal length [31, 37-39, 41, 42] and nasal depths (Ndepth1, Ndepth2) [31,37,38] have higher values in men than in women. Men are also characterized by higher values of the variable describing nasal dorsum convexity (Hump) [31,38]. On the other hand, no significant correlations were found between skeletal class and the size of the nasal dorsum hump [31,38]. Gulsen et al. [38] stated that NLA in class II is higher than in classes I and III.
Numerous authors [7-10, 28, 29, 31-34, 38, 43-47] have used NLA, enabling them to assess nasal position in the facial profile and, indirectly, the position of maxillary anterior teeth. NLA is very important for orthodontic treatment planning and is easy to measure. Table 5 shows that the values of NLA differ between ethnic groups: they are lowest in Korean and higher in European-American adults. NLA is similar in both sexes.
The following correlations were found between nasal structures and facial skeleton [38]: Ndepth1-positive correlation with mandibular length, posterior facial height, and hump; Nlength-positive correlation with anterior and posterior face height, maxillary and mandibular length, hump, and nasal bone length; Ndepth2-positive correlation with     Arshad et al. [31] reported the following findings: skeletal classes I, II, and III are characterized by different nasal profiles due to different values of NLA, soft tissue convexity, and low convexity of the nasal dorsum and significant differences exist between women and men concerning nasal profiles in terms of nasal length, nasal depth, hump, convexity columella, and nasal bone length.
In the studies by Gulsen et al. [38] and Arshad et al. [31], significantly higher values of the nasal length (Nlength), nasal depths 1 and 2 (Ndepth1 and Ndepth2), and nasal hump (Hump) were reported in men. The convexity of the lower part of the nasal dorsum (Cconv) and nasal bone length (NboneL) were higher in men in the study by Arshad et al. [31], whereas SFC was higher in men in the study by Gulsen et al. [38].
The results reported by Nehra and Sharma [34] indicate a significant correlation between nasal length and upper frontal facial height, inclination of the hard palate vault, and upper facial height. An upturned nose in adults is significantly correlated with maxillary anterior rotation [34]. This report is contrary to the study by Gulsen et al. [38], who found no significant correlation between the nasolabial angle (which is strongly correlated to the upturned nose) and facial skeletal parameters. Moreover, nasal length is correlated with palatal inclination NL (maxillary inclination) [34], similar to findings by Gulsen et al. [38], who observed an association between the nasal base angle and the inclination of the palatal plane.
Meng et al. [39] in their study on young Americans noticed that the proportion between upper and lower nasal height (3 : 1) is stable between the ages of 7-18 and is the same in both sexes. In males, continuous changes in nasal growth were observed between ages 7, 13, and 18. The nose grew more forwards than downwards [1,39,42]. Between the ages 13 and 18, the nose moved forwards more by the increasing nasal depth (Prn ′ -Prn) than by increasing the distance PMV-Prn ′ . The highest increase in females was noticed from age 7 to age 16. The nose grows more forwards than downwards, similarly as in men. In comparison, Buschang et al. [40] reported that in French-Canadian children, the nasal dorsum growth was upwards and forwards, but also in vertical growth, the nose directed downwards and backwards. These findings are consistent with the study by Robinson et al. [37], who noticed the presence of nasal hump in class II. However, in females, a lower increase in nasal depth was found than in males [11,39]. Meng et al. [39] concluded that nasal growth in males is still present after the age of 18. However, in females, the nose grows until the age of 16 [34,39,41,42].
Skinazi et al. [35] used Rickett's E-line and the Juanita Line. A sandwich is formed by these two lines and encloses the soft tissue profile in two thirds. Based on these measurements, Skinazi et al. [35] concluded that the mean upper lip, lower lip, chin, and total area were all statistically larger in men. They found no sexual dimorphism in the size of the nose [35]. Referring to nasal dimensions, similar results have been reported by Scavone et al. [8] (Japanese-Brazilian population), and contrary ones-by Anić-Milošević et al. [7] (investigation in Croatian and American population).   Convexity of lower part of nasal dorsum NboneL: Nasal bone length.

Data Availability
All data is a part of the manuscript.

Conflicts of Interest
The authors declare that they have no conflicts of interest.