Premium
Analyzing Facial Asymmetry in Children as a Function of BMI
Author(s) -
Bank Julia R.,
Wirawan Christian S.,
Ehrlich Daniel E.,
Marazita Mary L.,
Weinberg Seth M.,
Miller Steven F.
Publication year - 2020
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2020.34.s1.03726
Subject(s) - facial symmetry , fluctuating asymmetry , craniofacial , asymmetry , procrustes analysis , body shape , percentile , obesity , psychology , medicine , orthodontics , mathematics , statistics , biology , evolutionary biology , pathology , geometry , physics , quantum mechanics , psychiatry
Childhood obesity is a major public health concern in the US with a prevalence of nearly 19% for all children and adolescents. During growth, environmental factors such as poor diet can result in increased developmental stress. Given the symmetric patterning of facial growth in humans, any random deviations from symmetry, referred to as fluctuating asymmetry (FA), are considered a reflection of stress during growth. In this study, we examine whether childhood obesity serves as a source of developmental stress resulting in increased facial asymmetry. We hypothesize that a significant relationship between facial asymmetry and BMI exists, with higher BMI individuals demonstrating increased levels of facial asymmetry. To test our hypothesis, we obtained n=354 3D facial scans of children (age 3–19) from the University of Pittsburgh Center for Craniofacial and Dental Genetics. A total of n=286 facial scans represented children with a normal weight and n=68 faces were from children with obesity. Obesity status was determined using BMI values at or above the 95 th percentile. Facial scans were digitized using 34 coordinate landmarks to capture asymmetry in the eyes, nose, mouth, and lower jaw. Coordinate data was submitted to a Generalized Procrustes Analysis (GPA) to align the facial landmarks and geometric morphometrics (GM) was then employed to quantify facial shape phenotypes in MorphoJ. Principal Component Analysis (PCA) was used to identify patterns of facial asymmetry throughout the entire sample. Canonical Variate Analysis (CVA) was then employed to identify differences in facial shape due to BMI. Lastly, a Procrustes ANOVA was ran to quantify levels of fluctuating asymmetry in the sample and the resulting FA scores were regressed against BMI percentile to test for a significant relationship. The results from the PCA identified 43 principal components (PCs), with the first three PCs accounting for 44.3% of total facial asymmetry. PC1 (accounting for 20.6% of facial asymmetry) identified asymmetry in the position of the eyes and a slight deviation of the nose, mouth and lower jaw to the left or right. PC2 (12.9%) demonstrated asymmetry primarily between the jaw and nose. PC3 (10.7%) further illustrated asymmetries through the eyes and jaw. The results of the CVA showed a significant difference (p=0.023) in overall asymmetry between normal weight children with children with obesity. Normal weight children demonstrated a trend toward facial symmetry while obese children showed marked asymmetry with respect to the eyes, nose, and chin. Finally, the results of the regression between FA scores and BMI percentile did not produce any significant association (p=0.72). While these results indicate a significant difference in the patterning of facial asymmetry between normal and obese children, evidence for increased FA levels (and therefore developmental instability) was not observed. Although we reject our hypothesis, the results demonstrate interesting facial asymmetries between BMI groups that may be promising for future analysis with longitudinal data. Support or Funding Information Grant support: R01‐DE016148