Influence of genetic and environmental factors on transverse growth

Abstract

Objectives

The aim of the study was to determine the genetic and environmental effects on transverse growth of craniofacial structures, within and between identical and fraternal twins.

Methods

The sample consisted of 142 children in total, divided into 29 pairs of monozygotic twins, 42 pairs of dizygotic twins, and 1 set of dizygotic triplets. Postero-anterior cephalometric radiographs were taken at the ages of 9, 12, and 15 years. Intercanine width, maxillary width, mandibular width, nasal width, and facial width variables were measured. The genetic and environmental components of variance were analyzed with structural equation modeling for multilevel mixed effects.

Results

Intercanine width was initially mainly characterized by a moderate genetic component at 9 years (53%), with environmental influence increasing at age 12 (36%) and peaking at 15 years (84%). Maxillary width was under strong genetic influence at 9 years (70%), with genetic influence remaining strong up to 15 years (73%). Mandibular width was under additive genetic influence at 9 years (76%), with dominant genetic influence remaining high at 15 years (81%). Nasal width was under strong additive genetic influence at 9 years (69%) but switched to increased environmental influence at 15 years (59%). Finally, facial width had a moderate genetic influence at 9 years (66%), which increased at 15 years (90%).

Limitations

This study included patients of European descent, which may limit the generalizability of the findings to other ethnic groups.

Conclusions

Although monozygotic and dizygotic twins share at least part of their genetic material, environmental factors accounted for about 10%–84% of variability at various ages, with intercanine width being most affected.

Introduction

Craniofacial growth occurs in a specific completion sequence starting with the transverse dimension, followed by the sagittal then vertical growth [1]. Most transverse craniofacial growth is complete by age 18, with maxillary transverse growth peaking at 15 years of age [2, 3]. This phenomenon is essentially influenced by a sutural architecture, which divides the base of the skull, the maxilla, and the mandible into right and left parts, and involves a process of deposition and resorption that persists after sutural growth ceases [4, 5].

Many studies have looked into growth in the transverse plane, determining various factors that could influence craniofacial development [6–9]. Wagner and Chung found a relationship between transverse dimensions and facial patterns, noting that vertical growers tend to have less intermolar width gain [9]. Other studies highlighted the negative impact of habits such as mouth breathing and thumb sucking on transverse growth [10, 11]. Additionally, both environmental and genetic factors play a significant role in craniofacial growth, with approximately two-thirds of the human genes being involved in this process [12–14]. However, it remains unclear whether these factors have the same effect on the transverse dimension.

Growth evaluation is a critical component of orthodontic diagnosis and comprehensive treatment planning. In this regard, posterior anterior (PA) cephalometric radiography is commonly used to assess transverse dentofacial and/or skeletal discrepancies. Over the years, various cephalometric analyses of the transverse dimension, such as those by Ricketts and Grummons, have been developed to help clinicians identify the site of the discrepancy [15, 16]. Once the growth pattern is determined, specific orthodontic modalities can be recommended such as palatal expansion, orthognathic surgery, and/or functional devices [17].

Twin datasets are valuable in assessing the impact of genetic and environmental influences on growth and development. PA growth datasets of monozygotic (MZ) and dizygotic (DZ) twins/triplets are crucial assets in orthodontic diagnosis, treatment planning, prognosis, and growth prediction. Unlike static morphology analyses, evaluating the genetic factors on growth requires specific samples of age-matched siblings or twins, as repeated measurements on the same subject at different time points are essential for growth studies [18]. One of the most extensive growth datasets available for this purpose is the Forsyth Moorreess Twin Growth Study, which includes radiographs from 533 families of MZ and DZ twins. Although there have been few studies assessing the genetic and environmental components of growing identical and fraternal twins, these studies have mainly focused on lateral cephalometrics, evaluating anteroposterior and vertical facial growth [14, 19, 20].

To our knowledge, there is no longitudinal investigation that quantifies genetic and environmental influences on transverse facial growth of MZ and DZ twins. Therefore, the purpose of this retrospective cohort study was to assess the influence of genetic and environmental factors on transverse growth in untreated MZ and DZ twins.

Materials and methods

Study sample

The Institutional Review Board of Boston University (H-31945) granted ethical approval for the study. Subjects selected for this study were taken from the Forsyth Moorrees Twin Study, which includes annually collected records of 533 families of twins and/or triplets between 1959 and 1975. To determine zygosity, serologic testing was conducted on 29 factors, and additional indicators of phenotypic similarity were considered. All the subjects included in this study were of Caucasian origin and had no history of orthodontic treatment. The participants were selected based on the following inclusion criteria: (1) no history of craniofacial anomalies or chronic systemic disease, and (2) the availability of high-quality PA cephalograms for analysis. A total of 29 pairs of MZ twins, 42 pairs of DZ twins, and 1 set of DZ triplets were evaluated at the ages of 9, 12, and 15.

Cephalometric measurement

All PA cephalograms were captured using a standardized technique in centric occlusion, where the same device was used. The focus-coronal plane distance was 150 cm, and the film-coronal plane distance was 9 cm, maintaining a constant magnification factor of 6%. The patients were instructed to refrain from swallowing during the radiological examination. Tongue posture was then evaluated on the radiographs to confirm that no swallowing occurred during the procedure. The cephalograms were then scanned using an Epson Expression 11000XL scanner (Epson America, Inc, Long Beach, CA), set at a resolution of 300 dpi and 16-bit gray scale. The total number of PA cephalograms included in the data set was 992 radiographs, all in good quality.

Following the anonymization process, the scanned radiographs were traced by two individuals (SA and KC) by placing 10 cephalometric points (dental and skeletal) using OrthoDx^TM imaging tool. Ricketts analysis was used to derive 5 linear variables [15, 21].

In our study, the following variables were measured: intercanine width, maxillary width, mandibular width, nasal width, and facial width (Fig. 1).

• Nasal width: The distance between the widest points in the nasal cavity.

• Maxillary width: The distance between the right and left jugale points.

• Mandibular width: The distance between the right and left antegonial points.

• Facial width: The distance between right and left zygomatic arches.

• Intercanine width: The distance between the tips of the lower canines.

Statistical analysis

The descriptive statistics included the means and standard deviations, after checking for normality using the Shapiro-Wilk test. Crude differences between MZ and DZ twins were determined using generalized linear models, with standard errors clustered within twin/triplet families.

The influence of different factors on the observed outcomes was analyzed. These factors included additive genetic factors (A), non-additive (dominant) genetic factors (D) that include dominance and epistatic interactions between genetic loci, shared environmental factors (C) that contribute to variations among individuals within twin pairs but are not genetic, and unique environmental factors (E) specific to each individual. A pair of MZ twins possess an identical genetic composition, which indicates a genetic correlation of 1. The ACE and AE models consider genetic effects to be additive, meaning there is no interaction between genes. According to Mendelian inheritance rules, the genetic correlation within a pair of DZ twins is 0.5. ADE models consider interactions between genetic influences, resulting in nonadditive/dominant genetic effects. In these cases, the genetic correlation within a pair of DZ twins is 0.75, including a 0.5 additive genetic correlation and a 0.25 dominant genetic correlation. Other assumptions include the absence of assortative mating among the parents of twins and the assumption of equal environments for both MZ and DZ twins, while the AE/ADE models assume no environmental effects shared by both twins of a twin pair (no C).

The genetic and environmental components of variance were analyzed with structural equation modeling for multilevel mixed-effects ACE (additive-shared environmental-unique environmental) (Supplemental material 1). This modeling approach was implemented using Stata 14.2 (Stata Corporation, College Station, TX, USA). For each variable, an ACE, AE, and ADE model were fitted separately, with the letters representing the sources of variation allowed in each model. The Akaike information criterion (AIC) was used to discern the best-fitting model for all the variables, where a lower AIC value indicates a better fit of the model to the observed data. The same procedure was then applied to models adjusted for sex, and the best-fitting model was selected from the three. Finally, the best-fitting crude and sex-adjusted models were compared to identify the most parsimonious model, with the lowest AIC, which best explained the observed variance.

Finally, the narrow-sense estimate of heritability was computed for all factors by calculating the proportion of additive to phenotypic variance for all variables [22]. As adding patient sex in the structural equation model might lead to overfitting, a sensitivity analysis was performed by replacing all chosen models that incorporated sex effects with the next best-fitting model without any sex effects (Supplemental material 1).

The landmarks were plotted twice on 102 PA cephalograms by the same examiner (SA) one month apart for intra-observer reliability. A second investigator (KC) landmarked the same PA cephalograms to assess the inter-observer reliability. Method error was evaluated using the intra-class correlation coefficient (ICC) and the Bland-Altman limits of agreement.

Results

The repeated measurements analysis showed excellent intra and inter-examiner reliability, with small limits of agreement in all variables (Table 1). The study included 101 subjects (21 MZ twin pairs, 28 DZ twin pairs, 1 set of DZ triplets) at the age of 9, 109 participants (25 MZ twin pairs, 28 DZ twin pairs, 1 set of DZ triplets) at the age of 12, and 79 subjects (17 MZ twin pairs, 21 DZ twin pairs, 1 set of DZ triplets) at the age of 15 (Table 2).

In the descriptive analysis, no statistically significant differences were observed between MZ and DZ twins across all variables, except for the intercanine width at the age of 12 (P = .01) (Table 2).

Crude models were first constructed for all variance decompositions and then compared to models adjusting for differences according to sex (Table 3).

The ADE (additive-dominant-unique environmental) model better fit the following measurements: intercanine width (9 years), maxillary width (12 years), mandibular width (9 and 12 years), and nasal width (9 years). This indicated the presence of either nonadditive (dominant) or additive genetic influences for these measurements. The AE model better explained the nasal width (9 and 12 years). Models adjusted for sex were selected for some other measurements as shown in Table 4.

At the age of 9, the intercanine width showed a moderate additive and dominant genetic influence (53% and 37% respectively) and only a 10% environmental influence. It increased by 3.4 mm by the age of 12 years, with the unique environmental influence being slightly stronger (36%). From 12 to 15 years, there was no noticeable change in width, but the unique environmental influence became greater (84%).

At the age of 9, maxillary width was predominantly influenced by additive genetics (70%). It increased by 2.7 mm between 9 and 12 years, and another 2.7 mm between 12 and 15 years. The strong genetic influence persisted throughout this period, with a contribution of 73% at the age of 15 years. Similarly, mandibular width at 9 years was under strong additive genetic influence (76%). It increased by 4.5 mm between 9 and 12 years, followed by a further increase of 4.1 mm between 12 and 15 years. The dominant genetic influence on mandibular width remained significant at the age of 15 years (81%).

At the age of 9 years, nasal width was under strong additive genetic influence (69%), with similar increases in width between the ages of 9–12 and 12–15 years (1.6mm to 1.8mm). However, at 15 years of age, nasal width was under a stronger unique environmental influence (from 31%–59%).

Finally, facial width at the age of 9 years showed a strong additive genetic influence (66%), with almost similar increases in width between the ages of 9–12 and 12–15 years, ranging from 5.0 mm to 5.2 mm. At 15 years of age, the genetic influence on facial width further increased from 66% to 90% (Table 4, Fig. 2).

Moderate to high narrow-sense heritability estimates were observed for all the variables except for the intercanine width at 15 years, the maxillary width at 12 years, the mandibular width at 12 and 15 years, and the facial width at 12 years. The highest heritability was seen in the facial width at age 15.

Sensitivity analysis excluding sex effects was performed (Supplemental material 1) and indicated relative robustness. In all instances, the unique environmental variance was little to no different between the original and the sensitivity analysis. The only real difference was maxillary width at 15 years, where it was 73% additive in the original analysis and turned to 77% dominant in the sensitivity analysis.

Discussion

Several publications looked at the influence of genetics on dentofacial growth [18, 19, 23–26]. However, to the best of our knowledge, this retrospective cohort study is the first to quantify genetic and environmental influences on the transverse facial growth of MZ and DZ twins. In this study, 29 pairs of MZ twins, 42 pairs of DZ twins, and 1 set of DZ triplets were evaluated at the ages of 9, 12, and 15 years based on 5 transverse measurements on PA cephalograms.

According to the literature, the degree of heritability in craniofacial traits tends to increase with age [27]. This explains why in the present study, we investigated three time points—9, 12, and 15 years—to assess the genetic and environmental factors influencing craniofacial characteristics. Including both MZ and DZ twins allowed for increased accuracy in partitioning the genetic and environmental components of craniofacial variation, and the quantitative genetic modeling used to explore the heritability of transverse craniofacial morphology. However, as this study relied solely on radiographic data, further investigation through genome-wide association studies is necessary to assess the specific genes involved in craniofacial growth.

Drawing direct comparisons with other twin studies is challenging due to the rarity of similar investigations. In this research, 5 linear measurements were analyzed in 101 patients at the age of 9, 109 patients at the age of 12, and 79 patients at 15 years of age. Similarly to Nanda et al. results, all variables showed a constant increase during growth [3]. Additionally, all variables showed high concordance between twins, except for the intercanine width at age 12, indicating similar skeletal transverse relationships. This could be explained by their similar environment and genome [28]. However, narrow-sense heritability estimates varied among the measurements. All variables showed high heritability values at age 9 indicating a strong genetic component. However, all variables, except the nasal width, showed low heritability at age 12 or 15. For the intercanine width, there was a significant decline in heritability from 0.53 at age 9 to less than 0.01 at age 15, suggesting a stronger environmental influence with age. These findings contrast those of Eguchi et al. and Lin et al. who showed that the CE model was the most parsimonious model for the mandibular intercanine width [29, 30]. This could potentially be explained by the impact of environmental factors, such as function and habits, on mandibular teeth. Further research would study the influence of specific habits on the intercanine width of twins.

A similar pattern was observed for mandibular width, with heritability declining from 0.76 at age 9 to less than 0.01 at age 15. In contrast, maxillary and facial widths had strong heritability at age 9 with a sharp decline at age 12, followed by an increase to a high heritability at age 15.

The findings of the current study indicate that the ADE model better explained the observed variance for almost all variables, suggesting that both genetic factors and unique environmental factors contributed to most variables. Moreover, the AE model better explained the nasal width at ages 9 and 12, indicating that nasal width variance was due to both the additive genetic factors and non-shared environmental factors. Similar results were found in a study on the heritability of vertical craniofacial morphology in twins at 15 and 18 years of age [19]. This may be associated with airway obstruction, as inadequate nasal breathing can affect nasal and maxillofacial development contributing to atypical growth patterns.

Previous research noted the influence of sex on the heritability of craniofacial components [20]. Hence models adjusted for sex were selected for most variables, as they fitted the data better. At 9 years, models adjusted for sex were chosen for the maxillary width and facial width. At 12 years, models adjusted for sex were chosen for the intercanine width and the facial width, while at 15 years, they were chosen for the intercanine, maxillary, mandibular, nasal, and facial widths.

The results showed that most of the transverse variables are influenced by dominant and additive genetic factors, especially at the age of 9 and 12 years. This indicates that skeletal and dental transverse components are under significant genetic control and can be used as predictors of unfavorable transverse growth. For example, orthopedic palatal expansion is typically performed before the ossification of the mid-palatal suture. Therefore, the expectation of some relapse following palatal expansion could be explained by the strong genetic influence on maxillary width.

However, at 15 years of age, there was a dominance of factor E in the intercanine width (84%) and the nasal width (59%). The intercanine width displayed the most significant change, with additive genetic influence decreasing from 53% at age 9 to 16% at age 15, while the unique environmental contribution increased to 84%. These results were consistent with those of Lin et al. and Eguchi et al. [29, 30]. This could be explained by the cessation of increase in intercanine width once permanent dentition was reached, and the significant impact of environmental factors, such as habits and function [31]. Therefore, we could potentially conclude that intercanine width is more influenced by environmental factors at age 15. This could also imply that environmental factors are more involved in dental transverse dimensions [32]. Consequently, maintaining arch shape and controlling habits during orthodontic treatment is highly recommended to avoid relapse.

From a clinical perspective, a comprehensive understanding of the influence of genetic and environmental factors on the transverse dimension can help orthodontists better predict the response of skeletal and dental components to specific treatments at different ages. This can lead to more accurate diagnoses, customized treatment plans, and improved long-term stability.

Limitations

This study included both MZ and DZ twins/triplets, allowing for a more accurate partitioning of genetic and environmental components of transverse variables, However, due to the sample size, differences in statistical methods, and zygosity maturation, results should be interpreted with caution. Moreover, this study was purely clinical and radiographic, and genome-wide association studies are essential to determine the specific implication of any genes in transverse growth. Additionally, due to the lack of access to medical records, it was unknown if any of the participants had airway disorders or habits that could have potentially influenced the craniofacial growth. Furthermore, the patients were of European descent, which may limit the generalizability of the findings to other ethnic groups.

Conclusions

• Either additive or dominant genetic components were found at 9, 12, or 15 years of age for most transverse variables.

• Environmental factors accounted for 10%–84% of variability at different ages, with intercanine width being mostly influenced at 15 years of age.

• Significant heritability was observed in most variables. However, the heritability of maxillary, mandibular, and facial widths at age 12, along with intercanine and mandibular widths at age 15 demonstrated lower values.

Author contributions

Sameer Al-Obaidi (Investigation [equal], Methodology [equal], Writing—original draft [equal], Writing—review & editing [equal]), Spyridon Papageorgiou (Formal analysis [equal], Methodology [equal], Supervision [equal], Writing—review & editing [equal]), Marianne Saade (Formal analysis [equal], Writing—review & editing [equal]), Kristina Caradonna (Investigation [equal]), Alpdogan Kantarci (Resources [equal], Writing—review & editing [equal]), Leslie Will (Resources [equal], Writing—review & editing [equal]), and Melih Motro (Conceptualization [equal], Methodology [equal], Resources [equal], Supervision [equal], Writing—review & editing [equal])

Conflict of interest

The authors have no conflict of interest to declare.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References