Photogrammetric analyses of facial growth of Chinese in Hong Kong : traditional and geometric morphometric approach

Abstract

Introduction: Growth changes of facial features are influenced by changes in soft tissues and the underlying facial skeletons. Previous studies have mostly focused on skeletal changes with growth, while less attention has been paid to growth changes of soft tissue facial structures.

Objectives: This research project aimed to systematically review existing facial photogrammetric studies and to investigate growth changes of facial measurements and facial shape.

Methods: Longitudinal photogrammetric studies on age-related facial changes were examined in the first review. In the second review, Bayesian meta-analysis was performed to establish norms of facial measurements and to make comparisons among ethnicities/races. Based on a longitudinal population-representative sample of Chinese in Hong Kong, frontal and lateral facial photographs at age 12, 15, and 18 years were obtained. Traditional and geometric morphometric analyses were performed to evaluate changes of facial measurements and facial shape.

Results: The first systematic review identified only six eligible studies. These studies were methodologically heterogeneous and were subject to bias. The second review established ethnicity/race-specific norms of facial measurements and demonstrated that facial width and height, nasofrontal, nasolabial, and nasofacial angle differed among ethnicities/races.

Frontal and lateral facial photographs of 266 and 265 participants, respectively, were obtained at all three age levels. Using traditional photogrammetric analyses, significantly increased eye fissure height (p < 0.001) and decreased nasofrontal angle (p < 0.001) were noted in both genders. Vermilion-total upper lip height index increased in males (p = 0.020). Remarkable gender differences were noted for growth changes of nasofrontal angle (effect size: 0.55) and lower vermilion contour index (effect size: 0.59) from 12 to 18 years.

Geometric morphometric analyses showed that mean facial shape differed across the age levels (p = 0.001) in both genders. All pairwise comparisons were significant (p = 0.0001). Variance of facial shape decreased for frontal configurations of females and lateral configurations of both genders (p < 0.05). Gender differences were noted in orientation (p < 0.05) and shape (p < 0.05) of phenotypic growth trajectories. The amount of gender difference remained stable during growth.

When static allometry was taken into consideration, results of frontal facial shape analyses based on size-corrected Procrustes shape coordinates were very similar to the results obtained before size standardization. For lateral facial configurations, significant interaction between natural logarithm of centroid size and age by gender groups were noted (p = 0.0002). The groups’ static allometric trajectories were similar in rate of shape change (p > 0.05) but differed in directions (p < 0.05).

Conclusion: There is the distinct need for high quality longitudinal photogrammetric analyses of facial growth. Ethnicity/race-specific norms of facial measurements were established and inter-ethnic/racial facial variations were identified. Facial measurements and facial shape changed with growth, while the magnitude of measurement change and direction of shape change were different between genders. The influence of static allometry on analyses of growth changes of frontal facial shape during the observation period was negligible. Static allometric trajectories for lateral facial shape differed in directions, but were similar in rate of shape change.