A Study of Craniofacial Asymmetry in Monozygotic and Dizygotic Twins Using Posteroanterior Radiographs

Abstract

This project investigated craniofacial asymmetry in a prospective study among monozygotic (identical) versus dizygotic (fraternal) twins and sibling pairs. Posteroanterior radiographs that had previously been collected were used in this study and the zygosity of the twins had previously been determined. We hypothesized that in this study measuring craniofacial asymmetry there would be a greater concordance between monozygotic twins than between dizygotic twins for each of the variables measured. If there were some genetic component to asymmetry, we would expect to find a greater correlation between monozygotic versus dizygotic twins and sibling pairs since monozygotic twins are genetically identical and the dizygotic twins and siblings share on average fifty percent of their genes. If there are no statistically significant differences between the monozygotic and dizygotic twin correlations as shown from the heritability value, one could conclude that genetic factors are not important in the development of craniofacial asymmetry. This information is clinically significant. If we find that environment plays the predominant role in the development of asymmetry, one would conclude that clinicians should be vigilant about employing a treatment plan that will modify the environment while an asymmetry is developing. Since the zygomatico-frontal sutures and crista galli were found to be relatively symmetrical landmarks in a previous investigation by Harvold, a horizontal axis was constructed through the zygomatico-frontal sutures and a perpendicular vertical axis was constructed through crista galli. Perpendicular distances of ten bilateral structures were measured from the vertical reference line. Two methods of evaluating subjects' head positioning were used and compared. Any subject that was found to have excessive rotation of their head was excluded from the sample. After a preliminary analysis it was found that the horizontal distances were strongly correlated to patient positioning even after subjects determined to have excessive rotation of the head were eliminated from the sample. For this reason, vertical measurements of eight bilateral landmarks were made perpendicular to the reference line constructed through the zygomatico-frontal sutures. Using Statistical Package for the Social Sciences, the magnitude of craniofacial asymmetry was evaluated by calculating the absolute (non-signed) asymmetry values (IL-RI) for each of the ten horizontal and eight vertical variables. The horizontal and vertical variables were evaluated for any directional asymmetry to determine if the right or left side is consistently larger by calculating signed asymmetry values (L-R) for each of the ten horizontal and eight vertical variables. Since the subjects vary in size, relative differences were subsequently calculated. The relative difference was calculated by dividing the right-left absolute difference by half the sum of the right and left measurements [IL-Rlf.5(L+R)]. The relative difference of each of the ten horizontal and eight vertical variables was calculated for each individual. Using SAS software correlation coefficients were calculated for the relative asymmetry values of the monozygotic and dizygotic twin groups for each of the ten horizontal and eight vertical variables. Subsequently heritability values of craniofacial asymmetry were calculated for each of the eighteen variables using the standard formula h2= 2(rMz-rDz). The data revealed that head positioning had significantly affected our horizontal asymmetry values. We found a significant directionality in the horizontal and vertical dimension with the left side being larger horizontally and the right side being larger vertically. However the horizontal asymmetry values must be interpreted with caution as they are significantly affected by the rotation. After analyzing the data collected, the hypothesis was rejected. We were unable to demonstrate a significant difference in correlation values between monozygotic and dizygotic twins/sibling pairs for most craniofacial asymmetry values. The few variables that did show high heritability were in the horizontal plane. The results of the horizontal data must be interpreted with caution due to the significant effect of head rotation. In contrast it was found that vertical asymmetry values were not significantly correlated to rotation and none of these had notable estimates of heritability. These findings suggest that craniofacial asymmetry is more strongly influenced by environmental factors than genetics. This finding supports treatment by clinicians to improve or prevent developing asymmetries. While genetically controlled malformations can often be improved with treatment, environmentally induced malformations may respond well to treatment that changes the faulty developmental environment. Still it will be the capability of the patient to respond to the change in environment (treatment) that will determine the efficacy of the treatment. The interaction of genetic factors with this change in environment at a particular time as represented by the treatment may have a greater influence on the outcome of the treatment than how the problem was produced in the first place. Future studies analyzing asymmetry need to take extreme care in patient positioning and calibration of the x-ray machine, for they may lead to errors which overpower the data already subject to the error of a two-dimensional representation of a three dimensional object. Keeping these technical caveats in mind, future studies should focus on whether there are genetic factors that influence a patient's response (or the difference in response among patients) to the treatment of asymmetry.