INTRODUCTION

Although a wide range of environmental factors can exist, Class III malocclusion is thought to mainly develop by genetic influence. Schulze and Weise1 suggested a strong genetic role in mandibular prognathism (MP) in a twin study, which revealed that monozygotic twins have a sixfold higher chance of having MP than dizygotic twins.

There are several things to consider regarding the inherited susceptibility to MP in Class III malocclusion. First, the prevalence rate of MP is known to vary according to ethnicity. Hardy et al.,2 in a meta-analysis, reported that East Asians have the highest prevalence of MP, with approximately 18.0%, and are followed by Middle Easterners, Europeans, and South Asians, who have the lowest prevalence (1.2%). Second, skeletal Class III malocclusion can be divided into subtypes. Bui et al.3 classified skeletal Class III patients into five subtypes, which suggests that different genes might be involved in each subtype. Third, association between diverse genes and MP suggests that MP might be a complex disease. Recent advances in genetics and molecular biology have found several candidate loci for MP, such as 1p35–36, 6q25, 19p13, and 5p13 for Koreans and Japanese,4–8 1p36, 4p16, and 14q24–31 for Han Chinese,9–12 and 1p22, 3q26, 11q22, 12q13, and 12q23 for Hispanics.13 Some genes located within these loci are suspected to have a role in susceptibility to MP (perlecan, cartilage matrix protein 1, alkaline phosphatase, erythrocyte membrane protein band 4.1, growth hormone receptor, transforming growth factor beta 3, latent transforming growth factor beta binding protein 2, Ellis-van Creveld syndrome protein (EVC), EVC2, insulin-like growth factor 1, homeobox HOXC genes, and collagen type II alpha 1).4–13

Segregation analysis is useful for investigating the existence of major genes using the phenotype information gained from pedigrees of probands and for determining the best inheritance model that can explain the variation of phenotypes.14,15 Segregation analysis of MP has been conducted in only two ethnic groups: 37 Libyan families16 and 55 Brazilian families.17 These investigations suggested that MP has an autosomal dominant inheritance with or without incomplete penetrance. Interestingly, although Yamaguchi et al.5 reported that three markers (D1S234, D6S1689, and D19S566) were associated with MP in Koreans/Japanese, a follow-up study of Cruz et al.18 showed that these markers did not have any evidence of linkage in the Brazilian population. These findings imply that different genes might be involved in the etiology of skeletal Class III malocclusion for between Brazilian and Korean/Japanese.

It is necessary to perform a complex segregation analysis of Korean MP patients and their families, who belong to a homogeneous ethnic group and have the highest MP prevalence compared to other ethnic groups. Some authors of this study performed a preliminary study to estimate the affected ratio of the families and the heritability in Korean orthognathic patients and their families (100 probands, 2729 relatives; affected ratio, 3.5%; heritability (h²), 29.8%).19 However, it is still necessary to estimate familial correlations between possible pairs in the pedigree, to adjust sex and founder effect on heritability, and to perform segregation analysis. We also recruited additional pedigree members of the same probands to get more information from relatives and to increase the statistical power. Therefore, the purpose of this study was to investigate the existence of genetic influences on the epidemiologic incidence of MP in Korean Class III patients and their families using segregation analysis.

MATERIALS AND METHODS

Probands and Their Skeletal Characteristics

The probands of this study consisted of 100 skeletal Class III patients with MP (51 men and 49 women; mean age, 22.1 ± 5.2 years) who underwent orthognathic surgery at Samsung Medical Center in Seoul, Korea. All probands were diagnosed by cephalometric analysis as severe MP with no or mild maxillary retrusion (Table 1). Although the anteroposterior position of the maxilla (SNA, 81.2° ± 3.2°) and vertical pattern (FMA, 28.2° ± 6.2°) were within the Korean norm, the mandible was positioned forward (SNB, 84.1° ± 3.9°) beyond the Korean norm. The values of ANB (−2.9° ± 3.0°) and Wits appraisal (−11.7 ± 4.7 mm) revealed skeletal Class III relationships. The general or medical conditions that are related to MP or that might have affected the outcome of this study were excluded.

Table 1. Cephalometric Characteristics of Probands^a

View Fullscreen

Family Members

The participants included additional pedigree members of the same probands recruited in a previous preliminary study.19 As a result, 3777 relatives (1911 men and 1866 women) of the 100 probands belonging to at least three generations were included in the present study. Pedigree charts and questionnaires were designed to elicit information about probands and relatives (age, sex, and subtype of Class III malocclusion; Figure 1).19 Photographs and a clinical examination were used to determine whether each family member was affected with MP.

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

Familial Correlation

The intraclass correlation coefficient values of MP (Table 2) in full sibling pairs and in parent-offspring pairs were .2003 and .2036, respectively (all P < .001). However, the correlation estimate of MP incidence in spouses was not significant.

Table 2. Familial Correlations of the Mandibular Prognathism (MP) Between Possible Pairs of Family Members^a

View Fullscreen

Heritability

Although the ACE model fit better than the AE model (P < .05), the h² estimate of the ACE model (21.5%) was lower than that of the AE model (22.8%) after adjusting for sex and founder effects (Table 3).

Table 3. Heritability (h²) Estimates of the Mandibular Prognathism^a

View Fullscreen

One-Susceptibility-Type Models

The purposes of one-susceptibility-type analysis were to investigate whether covariates existed and which covariate should be included. According to the Akaike information criterion (AIC)21 value and the likelihood ratio test (Table 4), sex and founder status were included as covariates in subsequent segregation analyses.

Table 4. Parameter Estimates in the One-Susceptibility-Type Models With and Without Covariates^a

View Fullscreen

In addition, one-susceptibility-type analysis can determine which kind of familial association exists. When compared to various other familial associations, the AIC value of the three separate familial association coefficients (father-offspring, mother-offspring, sibling-sibling; δ_fo, δ_mo, δ_ss) was the lowest (1364.99; Table 5). Therefore, three separate familial associations were used in the subsequent segregation analyses.

Table 5. Parameter Estimate in the One-Susceptibility-Type Models, Incorporating Various Residual Familial Associations by the Multivariate Logistic Model multivariate logistic model^a

View Fullscreen

Two-Susceptibility-Type Models

The purpose of the two-susceptibility-type model was to investigate whether Mandelian dominant or recessive effects existed or not. All of the nonpolygenic multivariate logistic models (environmental-no-transmission, Mendelian dominant, and Mendelian recessive) and the finite polygenic mixed models (environmental-plus-polygenic, Mendelian dominant, and Mendelian recessive) were significantly different from the general model (all P < .001; Table 6). Since the general model showed a lower AIC than the other models, it can be regarded as a better fit model. These findings imply that the transmission of a potential major genetic effect for MP could not be confirmed.

Table 6. Analysis of MP with Two Susceptibility Types^a

View Fullscreen

Three-Susceptibility-Type Models

All of the nonpolygenic multivariate logistic models (environmental-no-transmission and Mendelian additive transmission) and the finite polygenic mixed models (environmental-plus-polygenic transmission and Mendelian additive transmission) were significantly different from the general model (all P < .001; Table 7). These findings indicate that the general model was a better fit (with a lower AIC) than the other models. However, when considering the AIC values, the Mendelian additive transmission was better than the environmental-no-transmission and the environmental-plus-polygenic transmission.

Table 7. Analysis of MP with Three Susceptibility Types^a

View Fullscreen

In conclusion, two- and three-susceptibility-type models showed that the general model was a better fit than the other models. Korean MP incidence did not have a major gene transmission model and was influenced by numerous minor effect genes and their additive effects.

DISCUSSION

The affected ratio of the relatives of Korean MP patients in this study was 5.3% (199/3777), which is similar to that in Lee et al.22 (4.5%; Table 8). However, these affected ratio values are lower than those of Japanese MP relatives (34.5% in Suzuki23 and 11.2% in Watanabe et al.24; Table 8). Although Koreans and Japanese are both East Asians, Koreans have lower affected ratios than Japanese. Since these Japanese studies had a relatively smaller number of relatives compared to probands than the present study (relatives vs probands: 1119 vs 243 for Suzuki,23 1480 vs 105 for Watanabe et al.,24 3777 vs 100 for this study; Table 8), this might account for the higher values of the affected ratio than that of Korean patients. In addition, Brazilian MP patients showed a 14.3% affected ratio (Table 8),17 which is higher than the 5.3% of this study. These results imply that different ethnic backgrounds can result in different affected ratios and prevalence of MP.

Table 8. Summary of the Results from Previous Studies on Class III Malocclusion^a

View Fullscreen

The estimated heritability of MP in our subjects (h²  =  21.5%) is lower than the 84.3% reported by Watanabe et al.24 and the 31.6% reported by Cruz el al.17 (Table 8), which suggests that MP might have a weak heritability in the families of Korean MP patients. Not only the difference in sample selection criteria and ratio of probands and relatives but also the variation of genetic traits and transmission pattern may affect the estimation of this value. In addition, since the relatives of MP patients might consider mild MP as normal, a more accurate classification method for MP and training of both investigators and participants are required for further studies.

Most of the complex diseases influenced by multiple genes can be classified into several subtypes. Various subtypes in skeletal Class III malocclusion occur because of the anteroposterior position of the maxilla, mandible, or combination of the maxilla and mandible, and differences in the vertical components.25 The ratios of these subtypes are expected to be different according to family and ethnic groups. Moreover, similar Class III malocclusion subtypes could have originated from different etiologies. The typical inheritance example of a certain Class III malocclusion subtype is the so-called “Habsburg Jaw.” Analysis of the skull of Joanna of Austria, a member of the Habsburg Royal family, revealed that a retrognathic maxilla rather than a prognathic mandible is a transmitted trait of this family.26 This finding suggests that different genes can be involved in the distinct subtypes of Class III malocclusion. Therefore, it is preferable to use the classified subtypes of Class III malocclusion to reveal the effects of genetic and environmental influences and their interactions on the cause of MP.

In segregation analysis of Class III malocclusion on 37 Libyan families, El-Gheriani et al.16 concluded that the autosomal-dominant model is the best fit. However, they did not differentiate the subtypes of Class III malocclusion (for example, MP and maxillary retrusion; Table 8). Therefore, their results could have originated from a mixture of Class III subtypes that might have different etiologic factors. In contrast, this study evaluated only the MP subtype of Class III malocclusion, and included more than 3.5 times the number of probands than El-Gheriani et al.16 (Tables 2 and 8). Nevertheless, we could not find any significant major gene inheritance model after analysis of 3777 relatives (Tables 4–7).

Cruz et al.17 also performed a segregation analysis on 55 Brazilian families with 2562 relatives. Their study was similar to this study in terms of the subject number and the recruited subtype of Class III malocclusion (only MP; Table 8). They reported that the autosomal dominant inheritance model with incomplete penetrance can explain the familial transmission of MP traits. However, they did not investigate three susceptibility type models and multiple additive polygenic effects. Therefore, the variation in inheritance patterns between their study and ours may be partly due to different statistical approaches.

According to the subsequent linkage analysis of Cruz et al.,18 the loci that are mapped to be connected with Korean/Japanese MP had no evidence of linkage with Brazilian families. Because they reported that Brazilian MP can be transmitted with the autosomal dominant mode in their previous study,17 this may suggest that Koreans and Brazilians have different genes, leading to different modes of transmission for MP, which caused the divergence in inheritance models.

The findings that the Korean population had a low affected ratio (5.3%) and relatively weak heritability (21.5%) of MP compared to other ethnic groups (Tables 2 and 8) and that no major gene transmission model was proven (Tables 4–7) might imply that the inheritance of MP in Korean Class III patients is likely influenced by numerous minor effect genes and their additive effects. Further studies are needed to verify these findings.

This study using segregation analysis of Korean MP patients and their families did not support the results from the studies of Libyans and Brazilians, which reported Mendelian autosomal dominant inheritance of MP.16,17 Increase in the sample size, establishment of the distinct subtypes in probands and relatives, comparison with different ethnic groups, and application of more sophisticated statistical methods may help to reveal genetic backgrounds underlying the etiology of skeletal Class III malocclusion more clearly. These investigations can provide more plausible models for inheritance and susceptibility genes of skeletal Class III malocclusion, and may help to find a prediction method, biologic marker, and cure for MP.

CONCLUSION

• The inherited susceptibility to MP in Korean Class III patients might be due to the summation of minor effects from a variety of different genes and/or influence of environmental factors, rather than Mendelian transmission of major genes.