Assessment of Cervical Vertebral Maturation and Chronological Age in Yemeni Children and Adolescents Using Lateral Cephalometric Radiographs

Introduction

Cervical vertebral maturation (CVM) serves as a biological marker for assessing skeletal maturation in patients undergoing orthodontic treatment. Chronological age (CA), which is often used as a primary indicator, has limitations due to individual variations in growth [1,2]. The timing of orthodontic treatment is critical for achieving optimal outcomes in skeletal and dental correction. Orthodontic interventions, particularly for skeletal discrepancies, are most effective during periods of rapid growth [1–4]. Predicting the peak of an individual’s growth potential, as this is when craniofacial changes are most likely to occur and is often referred to as pubertal growth spurt, is an essential aspect of modern orthodontic diagnosis and treatment planning [5]. While a retrognathic lower jaw in Class II skeletal cases is typically treated when mandibular growth is at its peak, the prognathic mandible in Class III skeletal patterns is preferably treated before this stage [4]. Conversely, surgical cases necessitate growth to finish [6].

Many biological indicators have been evaluated as possible markers of an individual’s peak growth. Secondary sexual characteristics and physical body measurements are considered appropriate indicators of skeletal maturation, but they cannot precisely predict the timing of maximum growth, due to their retrospective nature [7]. Alternatively, directly assessing skeletal maturation by observing specific bony maturation markers across various parts of the body is more valuable as a diagnostic tool for predicting the prospective adolescent growth spurt and for planning growth modification therapy [4,8]. More accurate methods that use radiographs have been detailed in earlier research, such as methods focusing on hand-wrist maturity [9,10] and the CVM stage [10,11].

CA has traditionally been used to recognize the stages of human ontogeny: child, young, or adult. Although CA is the easiest developmental age parameter to determine and can be considered as a predictor of growth [12], it is also an unreliable predictor due to significant inter-individual variations in growth patterns influenced by genetic, environmental, and hormonal factors [7,13–15]. Biological markers, such as skeletal maturity indicators, have been developed to address this limitation [16]. The CVM method has gained prominence due to its practicality and reliability [4,17–18]. By analyzing changes in the morphology of the second, third, and fourth cervical vertebrae (C2, C3, and C4), which are visible on lateral cephalometric radiographs (LCRs), CVM offers an accurate assessment of skeletal maturation, without additional radiographic exposure [19–22].

Previous studies across different populations confirmed the correlation between CVM stages and CA. For instance, in 2005, Baccetti et al established a 6-stage CVM method and demonstrated its efficacy in predicting pubertal growth spurt in European cohorts [4]. Baidas and Patil et al validated similar correlations in Saudi and Indian populations, respectively [14,23], noting sex-specific differences in growth patterns. Studies among same populations indicated a significant difference among sexes [24,25].

Initially, patients’ CA is used in diagnosis to assess the need for orthodontic treatment decisions. However, in cases that involve skeletal discrepancies, a more precise evaluation of skeletal maturation through CVM stages is essential. The estimates of age ranges for each CVM stage would be highly relevant clinically, as they could provide useful reference values for orthodontic diagnosis. In this study, we thoroughly assessed the age range linked to each CVM stage according to Baccetti’s method and explored how sex affects the age range of each CVM stage. We aimed to address the following questions: (1) “What CA corresponds to each of Baccetti’s CVM stages?” (2) “Is there a significant difference in CA based on sex for each of Baccetti’s CVM stages?” Therefore, this study aimed to evaluate the correlation between indices of CVM and CA in 240 LCRs of male and female Yemeni children and adolescents between 8 and 19 years of age.

Material and Methods

STUDY DESIGN, SETTING, AND ETHICAL CONSIDERATIONS:

An observational prospective lateral cephalometric study was conducted at the Faculty of Dentistry at Sana’a University, Yemen, for participants who were seeking orthodontic treatment. The participants were analyzed separately to account for differences in the timing of cervical vertebrae morphological changes. Ethical approval was granted by the review board university’s ethics committee with ref. #497 OB: 10/6/2024. Informed consent was signed by participants or guardians.

CEPHALOMETRIC SIZE CALCULATION:

The LCR size was calculated as mentioned in a previous study [26] using OpenEpi software 25 based on a 95% confidence interval, 85% power, and 0.5 margin of error. The required minimum LCR images were 110, for each sex, and 10 LCRs images were added to account for potential issues or defects during reading, ensuring data completeness and reliability. The included LCRs of 120 patients from each sex (total number 240) aged 8 to 19 years old were divided equally between male and female participants.

CEPHALOMETRIC CRITERIA:

The selection criteria included an LCR for a participant with Yemeni ethnicity; from both sexes; CA between 8 and 19 years; absence of systemic diseases or growth-related disorders; no history of cervical vertebrae trauma or surgery; and clear visualization of C2, C3, and C4 in LCRs. Patients with systemic growth disorders or poor-quality radiographs were excluded.

CEPHALOMETRIC SELECTIONS AND COLLECTION:

All LCRs had excellent clarity and good contrast, with a 1: 1 magnification, and were obtained from the same machine (PaX-Flex3D P2, Vatech, Korea). A selection process was used to ensure that the LCRs were evenly distributed across the required age range of 8 to 19 years. The identification and coding of LCRs were performed by the principal investigator (A. A. A.), who was responsible for the study objectives. Each participant was randomly assigned a numerical code.

CEPHALOMETRIC GROUPING AND ANALYSIS:

A total number of 240 LCRs were collected from patients who met the selection criteria. Each group was further categorized into 6 age subgroups (8–9, 10–11, 12–13, 14–15, 16–17, and 18–19 years), with 20 LCRs in each subgroup. LCRs were coded, and the following parameters were assessed and measured. CA was recorded as time from birth to time of the LCR examination date. For each patient, CA was registered in years and months and was obtained either by referring to the birth certificate or by directly asking patients or their parents. For CVM assessment, measurements were obtained from cervical vertebrae at each stage of CVM. CVM and classifications were adapted and verified according to Baccetti et al [17,27], as shown in Table 1.

Linear measurements were used to characterize the morphological changes of the lower borders of vertebrae C2, C3, and C4, as well as the body shapes of vertebrae C3 and C4. Points were connected to form planes, from which linear measurements were derived (Table 2, Figure 1) [1,25]. The linear measurements were used to measure the outcomes corresponding to various stages of CVM.

A schematic representation of the traced vertebrae demonstrating identified points and constructed lines is presented in Figure 2A and 2B to aid in evaluating distinct phases of CVM objectively.

Concavity depth of the inferior border is defined as the linear distance from deepest points of concavity depth to the created planes, which represent the lower borders of C2, C3, and C4 [17]. This distance was measured to the nearest 0.5 mm to determine the presence or absence of concavity (Figure 2B).

Vertebral body shape is identified to determine the CVM stages according to Baccetti et al [4]. The shape of the cervical vertebrae must be identified using an objective method described by Ishaq et al [27]. The ratio of line height and line width indicated the shape of the cervical vertebrae body, according to Ishaq et al [27]. The shape of the cervical vertebrae was classified according to the formation percentage into wedge-shaped, horizontally rectangular, almost square, square, and vertically rectangular, and counted as <65%, 65–80%, 80–95%, 95–105%, and >105%, respectively, as shown in Figure 2B.

Assessments of CVM stage were performed according to Baccetti et al and Safavi et al, with some modifications to CVM method (cervical stages [CS] from CS1 through CS6) [27,28] (Table 3). The overall flowchart of CVM assessment and measurement method for each LCR with their different classification stages, CS1 through CS6, is outlined in Figure 3.

INTRACALIBRATION RELIABILITY:

Intracalibration reliability was evaluated by using an intraclass correlation coefficient (ICC) from a reading of 20 LCRs, which were included among the total number of the LCRs assessed. All recorded values were higher than 0.91 (95% CI, 0.93–0.99). The LCR calibrations were conducted with a 10-day interval by a principal investigator (A. A. A.).

STATISTICAL ANALYSIS:

Statistical analysis was performed with IBM SPSS Statistics for Windows, version 26.0 (IBM Corp, Armonk, NY, USA). Inter-observer agreement was assessed using Cronbach’s alpha reliability coefficient and ICC. Descriptive statistics were generated by calculating the CA corresponding to the 6 stages of CVM indicators for an entire sample, as well as separately for male and female participants. An independent-samples t test was used to compare CVM between males and females. As the data were normally distributed, the Pearson correlation coefficient was used to determine the correlation between CA and CVM. Spearman rank correlation was also calculated, and the significance level was set at P≤0.05.

Results

DEMOGRAPHICS AND DESCRIPTIVE STATISTICS:

The mean CA of the study sample was 13.2±2.8 years. CVM stages were higher at CS5 (fifth cervical stage), followed by CS4 (fourth cervical stage) with 74 (30.8%) and 64 (26.7%), respectively, while the lowest was for CS3 (third cervical stage) with 18 (7.5%), as shown in Figure 4.

CORRELATION BETWEEN CA AND CVM:

The Spearman rank correlation demonstrated a statistically significant relationship between CA and CVM stages for both sexes (r=0.887; P<0.05). Sex-specific analysis revealed earlier attainment of CVM stages in females than in males. For example, CVM Stage III was between 10.25 and 13.18 years in males and 10.04 and 12.15 years in females (Table 4).

Table 5 and Figure 5 show a correlation between CVM stage and CA in male and female participants. Statistical analysis using an independent-samples t test showed statistically significant differences between males and females in the CS4 and CS5 stages (P<0.05). No significant differences were found in the rest of the CVM stages.

Discussion

Maturation is a key concept for orthodontists when evaluating a growing child, particularly those with dentofacial issues. Several researchers studied various maturation indicators, such as CA, hand-wrist ossification, CVM, and dental maturation, to determine if a correlation exists between skeletal maturation and these parameters [4,29–31]. Previous reports indicated that CA is not a reliable predictor of pubertal growth spurt, because of significant variations in timing among individuals [9,32]. In the present study, we evaluated the correlation between CVM stage and the average CA in a Yemeni sample. The null hypothesis was rejected because a significant difference was documented between sexes and maturation of CVM stages and CA.

CVM was selected from various indicators of skeletal maturation because it is considered highly reliable. Originally proposed by Lamparski in 1972 [33] and later refined by Hassel and Farman in 1995 [32], this method was improved further through successive studies, as noted earlier [4,11,18]. Baccetti et al adapted the original Hassel and Farman CVM method [32]. This method was chosen for the present study because of its extensive use in existing literature and because it is easier to administer than other methods, given that the 2 anatomical variables in each stage simplify the classification process. Despite being relatively new, this method has proven its validity and has been used in studies aimed at developing a new formula for assessing the potential for mandibular growth [14,34,35].

The age of participants in this study ranged from 8 to 19 years, intentionally including both the early and late onset of skeletal growth, which are important considerations for orthodontists. The pubertal growth spurt typically occurs between the ages of 10 and 15 years [7,9]. This age range is similar to that used in a study on a Chinese population [35]. Meanwhile, studies conducted on a narrow age range from 8 to 14 years [36] and 5 to 16 years [37] concluded that maturation stages CS5 and CS6 were not observed.

The findings of the present study showed a strong correlation between CA and CVM for each sex, with correlation coefficients of r=0.898 for boys and r=0.890 for girls, and r=0.887 overall (P<0.001). This finding suggests that CA might generally serve as a suitable gauge for initial assessment of skeletal maturation. However, a crucial detail to note is the significant individual variability in the onset and deceleration of the pubertal growth spurt (CS3 and CS4). A clear sex-based difference in skeletal maturity was observed. In Yemeni females, CVM stages were linked with ages ranging from 8.29 to 17.77 years, while in males, the range was from 8.83 to 18.36 years. This finding indicates that Yemeni females generally mature earlier than males and aligns with previous studies conducted in various populations using similar methodologies [38]. It is also consistent with studies that used hand-wrist maturity to evaluate skeletal maturation, which found that females generally reach maturity at a younger average age [30,39]. Baccetti et al and Baidas reported similar sex-based maturation differences, with females reaching skeletal maturity 1.5 to 2 years earlier than males [4,14]. The findings of this study also agree with those of Perinetti et al, which demonstrated an earlier pubertal growth spurt in females across multiple populations [10]. This could be related to hormonal factors among females.

Baidas found a relatively strong correlation between CA and skeletal age assessment, with a correlation coefficient of r=0.864 in Saudi adolescents [14], which aligns with the present study’s findings of a strong correlation between CA and CVM for each sex, with an overall coefficient of r=0.887, and r=0.898 for boys and r=0.890 for girls. A study conducted in a sample of 400 Chinese participants in 2008 reported a slightly lower correlation between CVM and CA than that in the present study, with correlations of r=0.757 for boys and r=0.787 for girls [40]. The reason for this could be a genetic and environmental factors.

One potential explanation could be the variance in age distribution between samples. In the study by Alkhal et al, the age of female participants ranged from 10 to 15 years, and that of male participants ranged from 12 to 17 years, whereas the sample of the present study included participants aged 8 to 19 years for both sexes [40]. The broader age range in this study sample may explain the higher correlations that were observed, as individual differences in growth and development, including skeletal age, tend to increase with age. Ethnicity might also play a role, as growth patterns in Chinese populations may differ somewhat from those in individuals from Yemen [41].

In contrast, weak correlations were found between CA and CVM in females (r=0.5471), a strong correlation in males (r=0.6451), and a moderate correlation in the overall sample (r=0.5631) [42]. In a different study, weak correlations were observed between CA and CVM in a central Indian population for both sexes (r=0.5096); the sample size was 125 LCRs, and the age ranged between 10 and 15 years [23]. These findings contrast with the results in the present study. This difference may be due to a small sample size and a narrow age range or to racial and sex differences in the timing of maturation [41,43].

Overall, we can conclude that reasons for over- or under-estimation of CVM by CA can be attributed to several biological and environmental factors. One important reason is the variability in the onset of puberty, which differs significantly across individuals due to genetic predisposition [21]. On the other hand, early maturing individuals can show advanced CVM stages at a relatively younger age [6,7,37].

CVM was assessed on the basis of its 6 stages, as defined earlier [4]. This approach used the second, third, and fourth cervical vertebrae, all of which were visible in all LCRs. This method has been recommended and used in various recent studies [10,17,25]. In the present study, several cephalometric measurements were used to accurately assess the shape changes of the cervical vertebrae. First, the height-to-width ratio of the cervical vertebra body was determined, assigning values that categorize its shape as trapezoidal, horizontally rectangular, square, or vertically rectangular. Second, the concavity of the inferior border was quantified using a linear measurement rather than a subjective visual assessment, classifying it as absent or present. These measurements enhanced the consistency and precision of the observations. This methodology builds upon the foundation of previous researchers [17,27,38,44].

Our results demonstrated a statistically significant correlation (r=0.887, P<0.000) between CVM stages and CA, affirming an efficacy of CVM as a skeletal maturity indicator. This finding aligns with previous studies conducted in different populations, including those that reported similar findings in European and Saudi cohorts [4,14]. Baccetti et al found a strong correlation (r>0.8) in Italian adolescents [4], while Baidas reported a slightly lower correlation in Saudi patients [14]. Similarly, research conducted among Indian and Korean samples indicated a significant correlation, reinforcing the reliability of CVM as a predictor of skeletal maturity across diverse populations [24,45]. A minor difference observed in correlation strength may be attributed to ethnic and genetic variations.

The present study results indicated that CVM Stages III and IV represent a peak period for pubertal growth, making them an ideal time for orthodontic intervention. Functional appliances are most effective when introduced during these stages, to maximize skeletal correction potential. This observation is supported by a group of studies conducted by Baccetti et al [4,11], which reported that the peak mandibular growth occurs around CVM Stages III and IV. Similar trends were observed in research on orthodontic interventions in Saudi Arabian [14] and Indian [24] populations, reinforcing the idea that the pubertal growth spurt is the most critical period for initiating orthopedic treatment.

The timing of CVM stages in a Yemeni sample was compared with data from other populations, including studies conducted in Saudi Arabia, India, and Europe. While a general pattern of skeletal maturation was similar, slight variations were noted, possibly due to genetic, nutritional, and environmental factors. For example, Alkhal et al and Baccetti et al reported that the mean CA for CVM Stage III in European and Middle Eastern populations ranged from 11.2 to 13.5 years [4,11,40], which is slightly higher than in a Yemeni cohort. These differences suggest that ethnicity and environmental influences, such as diet and socioeconomic factors, play a role in skeletal growth timing.

The findings of the present study align with results of a study that found high correlation values between CA and CVM among Indian schoolchildren, ranging from r=0.779 to r=0.748, while the mean CA for males with CVM was 8.5 years for CS1, 8.5–10.5 years for CS2, 11.5–13.5 years for CS3, and 13.5 years for CS4. For females, CS1 was absent, and the mean CA was 8.5–9.5 years for CS2, 10.5–11.5 years for CS3, 12.5 years for CS4, and 13.5 years for CS5 [36]. Compared with the findings of the present study, earlier findings may reflect racial or environmental factors [41], and an absence of CS6 could be attributed to narrow age range of 8 to 14 years. In a study conducted at Istanbul Medipol University, strong correlations were identified between CA and CVM assessment among participants aged 7 to 18 years. The correlation coefficients were 0.744 for males and 0.778 for females [46]. These findings align with those of the present study, and the reasons may be attributed to similar socioeconomic and nutritional statuses between the 2 populations.

The visible concavity on the lower border of C3 indicates the stage just before peak growth [4]. In the present study, this stage occurred at an average age of 10.25±1.500 years in boys and 10.04±1.035 years in girls. CS3 is an ideal time to start functional jaw orthopedics, as peak mandibular growth occurs between CS3 and CS4 [4]. These results show that the CVM can be considered a more accurate indicator of pubertal growth spurt in Yemeni individuals than can CA. Thus, this method is used as a secondary indicator for orthodontic diagnostics in Yemeni patients.

The findings of this study revealed a statistically significant correlation between CA and CVM stages. However, it is important to note that, as established in the literature, CA is not a direct determinant of skeletal maturity, due to individual variability in growth patterns. Therefore, although a strong correlation exists, it does not imply that CA can accurately predict an individual’s position on the growth curve or their proximity to the peak growth spurt. However, these results support the use of CVM stage as a reliable marker of skeletal maturation, with CA serving only as an approximate reference. This correlation can be useful for age estimation in a context of physical anthropology and forensic science, rather than as a sole diagnostic tool for growth-based orthodontic treatment planning.

We measured the CVM stage by using LCRs from participants in Yemeni; this can be considered a strength of this survey. However, one limitation is the relatively small number of LCRs analyzed, all of which were collected from a single city. Potential examiner bias due to the use of a single observer may be considered a limitation, along with variations in image quality. Additionally, hormonal changes associated with increasing chronological age, which can affect bone maturation, were not assessed. Further research is needed with a larger sample size and sex distribution, to establish specific maturity standards for the Yemeni population. We also suggest conducting a study to determine the relationship between CVM and skeletal maturity by examining additional indicators such as sternal fusion, ossification of the sternal end of the clavicle, and epiphyseal ossification of the long bones.

Conclusions

A significant correlation was found between CVM stages and CA among growing Yemeni individuals, with females attaining skeletal maturity earlier than males. These findings provide a guide for optimizing orthodontic treatment timing in Yemeni populations. Future research should explore using longitudinal designs, include LCRs from different dental centers and cities, and incorporate additional biological markers as well as new digital programs to enhance predictive accuracy in orthodontic planning.