Received: 28 August 2020 | Revised: 10 October 2020 | Accepted: 4 December 2020

DOI: 10.1111/ocr.12457

ORIGINAL ARTICLE

Predicting mandibular growth based on CVM stage and gender

and with chronological age as a curvilinear variable

Lorenzo Franchi1,2 | Michele Nieri1 | James A. McNamara Jr2,3,4,5 |

Veronica Giuntini1

1
Department of Experimental and Clinical
Medicine, Orthodontics, The University of Abstract
Florence, Florence, Italy Objective: The aim of this study was to develop a prediction model that combines
2
Department of Orthodontics and Pediatric
the information derived from chronological age (analysed as a curvilinear variable),
Dentistry, School of Dentistry, The
University of Michigan, Ann Arbor, MI, USA gender and the CVM method to predict mandibular growth.
Settings and Sample Population: 50 participants (29 females, 21 males) were se-
3
School of Medicine, The University of
Michigan, Ann Arbor, MI, USA
4
lected from the AAOF Craniofacial Growth Legacy Collection, the Michigan Growth
Center for Human Growth and
Development, The University of Michigan, Study and the Denver Child Growth study.
Ann Arbor, MI, USA Materials and Methods: In this investigation, 456 lateral cephalograms were ana-
5
Private Practice of Orthodontics, Ann
lysed by applying a mixed effect model. The outcome variable was the annualized
Arbor, MI, USA
increment in total mandibular length (Co-Gn) during the year following the lateral
Correspondence
cephalogram on which the cervical stage and chronological age were evaluated. The
Lorenzo Franchi, Department of
Experimental and Clinical Medicine, School predictive variables were chronological age up to the fifth order, gender, stage of
of Dentistry, Università degli Studi di
cervical vertebral maturation, as well as interactions between age and gender, age
Firenze, Via del Ponte di Mezzo, 46-48,
50127 Firenze, Italy. and cervical stage, and gender and cervical stage.
Email: [email protected]
Results: Cervical stage, chronological age up to the fourth order, gender, and the
interaction between age and gender were significant predictors of annualized incre-
ments in mandibular length. The annualized increment in Co-Gn was significantly
greater for CS 3 when compared to all other cervical stages. Further, annualized in-
crements in Co-Gn for CS 1 and CS 2 were significantly greater when compared to
CS 5.
Conclusions: Cervical stage, chronological age and gender can be used jointly to pre-
dict the annualized increment in mandibular growth. Cervical stage 3 exhibited the
greatest annualized increase in mandibular length.

1 | I NTRO D U C TI O N Class II patients with mandibular deficiency is during the pubertal

phase of development.1-3 Cephalometric investigations on longitu-
The issue of optimal timing in dentofacial orthopedics is linked in- dinal samples have identified a pubertal spurt in mandibular growth
timately to the identification of periods of favourable growth that that is characterized by large individual variations in onset, duration
can contribute to the efficient and effective improvement of skeletal and rate.4-7
problems in the individual patient. For example, it has been shown Several indicators of individual skeletal maturity have been pro-
that the best timing to stimulate mandibular growth effectively in posed over the years to define treatment timing in orthodontics.

© 2020 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

Orthod Craniofac Res. 2020;00:1–7.  wileyonlinelibrary.com/journal/ocr | 1

2 | FRANCHI et al.

The most commonly used indicators of individual skeletal matu- • Availability of a series of at least 6 consecutive annual lateral
8,9
rity are increase in statural height, skeletal maturation of the cephalograms from the age of 7 to 18 years;
hand and wrist,10,11 and the maturation of the cervical vertebrae • The bodies of the second, third and fourth vertebrae had to be
(CVM method).12-14 Several studies have investigated the role of visible in all films;
the CVM method in predicting the pubertal growth spurt in the • The interval between 2 consecutive cephalograms had to range
mandible.9,15-25 from a minimum of 6 months to a maximum of 18 months;
To our knowledge, however, there is lack of information on the • The first cephalogram of the series had to show CVM stages CS 1
predictive power of the CVM method on mandibular growth in or CS 27,8; and,
growing subjects. 21,22 Thus, the aim of the present study was to de- • The last cephalogram of the series had to show at least CVM stage
velop a predictive model that combines the information derived from CS 5.7,8
chronological age, gender and the CVM method to predict mandib-
ular growth. In this investigation for the first time to our knowledge, Exclusion criteria were incomplete records, radiographs of poor
chronological age was used as a curvilinear variable. quality, anomalies in vertebral morphology and evidence of ortho-
dontic treatment (except when passive space maintainers were ev-
ident in the cephalograms). Two operators selected independently
2 | M ATE R I A L S A N D M E TH O DS the participants from the Growth Studies. A third operator was con-
sulted in case of uncertainties.
Participants who had not received orthodontic treatment were Increases in mandibular length (Co-Gn, the linear distance from
selected from the records of the Fels, Iowa, Mathews and Oregon Condylion to Gnathion) between consecutive cephalograms taken
Growth Studies that are available through the American Association annually were calculated for the entire series of cephalograms for
of Orthodontists Foundation (AAOF) Craniofacial Growth Legacy each participant. In that the interval between consecutive cepha-
Collection Project (www.aaofl​egacy​colle​ction.org). Moreover, the lograms was not always 12 months, the increases in Co-Gn were
complete records of the University of Michigan Growth Study (after annualized. The outcome variable was the annualized increment in
having removed the 30 subjects that were used in the previous study mandibular length during the year following the lateral cephalogram
by Baccetti et al13 to elaborate the CVM method) and of the Denver on which the cervical stage and chronological age were determined.
Child Growth Study were screened. Mandibular length was measured on all cephalograms by the
Inclusion criteria were as follows: same examiner (LF) on the digital cephalograms using a cephalometric

F I G U R E 1 In the prediction model, chronologic age was considered as a polynomial up to the fifth order or fifth degree. Graphical
representations of polynomials. A, First order or first degree polynomial (linear polynomial or linear function). B, Second order or quadratic
term or second degree polynomial is represented graphically by a parabolic curve. C, Third order or cubic term or third-degree polynomial
has a single inflection point, (a point where the function changes from being concave to convex or vice versa) and 2 extrema (maximum or
minimum). D, Fourth order or quartic term or fourth degree polynomial has 2 inflection points and 3 extrema. E, Fifth order or quintic term
or fifth degree polynomial has 3 inflection points and 4 extrema
FRANCHI et al. | 3

software (Viewbox 4.0, dHal Software). The value of Co-Gn was The level of statistical significance was set at P < .05. The sta-
standardized to 0% enlargement (life size) after adjustment of the tistical computations for the intra-examiner reproducibility and for
magnification factor of the different Growth Studies. Point Gnathion the mixed model were performed with specific software (MedCalc®,
was defined as the most anteroinferior point on the contour of the version 19.0.3, MedCalc Software bvba and JMP® vers. 13.0.0, SAS
bony chin. It was determined by bisecting the angle formed by the Institute Inc).
mandibular and facial (Nasion-Pogonion) planes.
The CVM method used in the present study was the one de-
scribed by McNamara and Franchi.8 The intermediate or in-between 3 | R E S U LT S
stages (when the characteristics of two consecutive stages were
present in a single image) 8 were included in the more immature From a parent sample of 1151 subjects, a final sample of 50 par-
stage, that is the intermediate CS 3-4 was classified as CS 3. ticipants (29 females and 21 males) was derived. Six cases (5 fe-
The predictors used in developing the multivariable prediction males and 1 male) were derived from the Fels Growth Study, 2 (1
model first was chronological age up to the fifth order (the second
order, or quadratic term, describes a parabolic curve while the third TA B L E 1 Frequencies and percentages of the CVM stages with
correspondent mean and standard deviation of the annualized
order, or cubic term, indicates the presence of inflection point in the
increment in Co-Gn
curve; in general, the higher the order, the more complex the poly-
nomial curve; Figure 1). Then, gender was considered. The stage of Annualized
development of the cervical vertebrae also was determined as cat- increment in
CoGn (mm)
egorical variable (CS 1, CS 2, CS 3, CS 4, CS 5 or CS 6). Interactions Frequency
among these variables also were evaluated. All cephalograms were Vertebral stages (n) Percentage Mean SD
staged according to the CVM method by a single expert examiner (LF). CS 1 140 31% 2.3 0.9
All participants from the Growth Studies who met all inclusion/ CS 2 62 14% 2.4 0.9
exclusion criteria were included in this study. We followed the
CS 3 86 19% 3.5 2.0
TRIPOD statement 26 for transparent reporting of a multivariable
CS 4 57 12% 1.8 1.5
prediction model. This study was conceived as a Type 1a analysis26
CS 5 72 16% 1.0 1.0
(development of a prediction model where predictive performance
CS 6 39 9% 1.1 0.9
is evaluated directly using the same data).
Total 456 100% 2.3 1.5

2.1 | Statistical analysis TA B L E 2 Mixed effect model

Intra-examiner reproducibility for the CVM stages and for the Variable Estimate SE P value

Co-Gn measurement was calculated on 30 randomly selected ceph- Constant (intercept) 64.9905 25.54
alograms after 2-week washout period with the weighted kappa CS1 0.9211 0.43 <.0001
coefficient for ordinal data and with the intraclass correlation coef- CS2 0.9660 0.44
ficient, respectively. Random error for the Co-Gn measurement was CS3 1.8190 0.39
assessed with Dahlberg's formula.
CS4 0.3245 0.37
Descriptive statistics were performed for the following vari-
CS5 −0.0984 0.37
ables: gender, chronological age and CVM stage.
Age (y) −21.2229 8.55 .0135
As for inferential statistics, a mixed effect model was applied. As
Age2 2.5398 1.05 .0161
mentioner earlier, the outcome variable was the annualized increment
Age3 −0.1286 0.06 .0223
in mandibular length (Co-Gn) during the year following the lateral
cephalogram on which the cervical stage and age were determined. Age 4 0.00233 0.001 .0347

Random effect was the participant (random effect is the effect that Gender F −0.6640 0.12 <.0001

arises from uncontrollable variability within the sample, and it is usu- (Age-11.87) × gender (F) −0.1556 0.05 .0037
ally attributed to the participants). The predictive variables (fixed Note: Outcome variable: annualized increment in CoGn. Predictive
effects) were chronologic age up to the fifth order, gender, cervical variables: CVM stage, age up to the fourth order, gender and interaction
stages (categorical variable—CS 1, CS 2, CS 3, CS 4, CS 5 or CS 6), in- age × gender. In presence of CS6, the estimate is 0. For male subjects,
the estimate is 0. Age5 was not statistically significant.
teractions between age and gender, age and cervical stage, and gender
Formula for CS3 Female subject at 12 y: 64.9905 + 1.81890 −
and cervical stage. The interactions and age from second up to fifth
21.2229 × 12 + 2.5398 × 122 − 0.1286 × 123 + 0.00233 × 124 − 0.6640
order were included in the model only if they were statistically signif- − 0.1556 (12 − 11.8714) = 3.27
icant. Post hoc test for CVM stages was evaluated with Tukey HSD. A Formula for CS2 Male subject at 10 y: 64.9905 + 0.9660 −
residual graphical analysis was performed to test model assumptions. 21.2229 × 10 + 2.5398 × 102 − 0.1286 × 103 + 0.00233 × 10 4 = 2.41
4 | FRANCHI et al.

F I G U R E 2 Annualized increments of
Co-Gn (y axis) as of function of age (x axis),
gender and cervical stage

TA B L E 3 Statistical comparison between the cervical stages As for inferential statistics (Table 2), the model showed a signifi-
cant effect for cervical stage, for age up to the fourth order, for gen-
Adjusted Diff from 95% CI Diff
CVM mean (mm) CS3 (mm) from CS3 Diff der and for the interaction between age and gender. Cervical stage
CS 3 exhibited the greatest annualized increase in Co-Gn. The fact
CS1 2.4 0.9 0.2; 1.6 B
that chronologic age was significant up to the fourth order implied
CS2 2.4 0.9 0.2; 1.5 B
that the annualized increase in Co-Gn varied in a complex manner as
CS3 3.3 – – A
a function of age (see the complex curves in Figure 2).
CS4 1.8 1.5 0.8; 2.2 BC
As for gender, the annualized increase in Co-Gn was significantly
CS5 1.4 1.9 1.1; 2.7 C greater in males. The significant interaction between age and gender
CS6 1.5 1.8 0.7; 2.9 BC indicated, however, that the difference between males and females
Note: The mean of the annualized increment in Co-Gn was adjusted for varied with age. As chronological age increases, the difference be-
age and gender. The column ‘Diff from CS3’ reports the differences tween males and females also increased, favouring males. All other
between the CS3 mean and the means for the other CVM stages. The
interactions and the fifth order for chronologic age were not signif-
column ‘Diff’ indicates the statistically significant differences between
icant and, therefore, not included in the final model. The analysis of
the stages. In the last column, stages with a different letter indicate a
statistically significant difference. residuals did not show any deviation from assumptions of the model.
Abbreviation: Diff, Difference. Figure 2 was constructed based on the estimates of the mixed ef-
fects model reported in Table 2. Two examples of how to derive two
points for the curves in Figure 2 are reported at the bottom of Table 2.
female and 1 male) from the Iowa Growth Study, 3 (1 female and 2 Table 3 reports the comparisons for the annualized increments in
males) from the Mathews Growth Study, 24 (15 females and 9 males) Co-Gn for each cervical stage. The annualized means were adjusted
from the Oregon Growth Study, 3 (1 female and 2 males) from the for age and gender. The annualized increment in Co-Gn was signifi-
Denver Child Growth Study and 12 (6 females and 6 males) from the cantly greater for CS 3 when compared to all other cervical stages.
University of Michigan Growth Study. Also, the annualized increments in Co-Gn for CS 1 and CS 2 were
The intra-observer reproducibility for the CVM method and for significantly greater when compared to CS 5.
27
the Co-Gn measurement was ‘almost perfect’ (CVM method: 0.87,
95% confidence interval, CI, 0.77-0.96; Co-Gn: 0.99, 95% CI 0.99-
1.00). The random error for the Co-Gn measurement was 0.57 mm. 4 | D I S CU S S I O N
In this investigation, 456 lateral cephalograms were analysed.
The mean number of cephalograms per participant was 9.1 ± 1.2 The aim of the present study was to develop a predictive model for
(minimum 6 and maximum 12 cephalograms). In Table 1, the frequen- mandibular growth using chronological age, gender and the CVM
cies of the different CVM stages and the corresponding percentages method as predictive variables.
and annualized increments in Co-Gn are reported. The mean age at A unique feature of this study was entering chronological age in
the first cephalogram was 8.2 ± 0.5 years (min 7.5, max 9.9 years) the predictive model as a curvilinear variable (polynomial curve up to
while the mean age at the last cephalogram was 16.5 ± 1.1 years (min the fifth order or degree, Figure 1). Using age as a curvilinear variable is
14.0, max 20.2 years). The mean interval between 2 consecutive important when analysing mandibular growth changes with age. If age
cephalograms was 1.0 ± 0.1 years [min 0.75, max 1.49]. is entered as a linear variable (Figure 1A), this means that we expect
FRANCHI et al. | 5

that mandibular growth should increase or decrease linearly along with the CVM method in predicting mandibular growth in female subjects
age. We know, however, that mandibular growth is not linear with age with Class II malocclusion selected from the Nijmegen Growth Study.
but rather follow a curvilinear trend that is characterized, particularly A linear mixed model was applied in order to determine potential as-
during adolescence, by an acceleration that reaches a peak, followed by sociations between increments mandibular growth (evaluated in a
a deceleration in mandibular growth rate until the end of active growth. 6-month interval) and average across observers CVM scoring after
In the literature, there is controversy on the reproducibility of the adjusting for age.
CVM method assessment. Some studies have reported poor reproduc- A major limitation of the above-mentioned studies is that the
ibility of the CVM method,28 whereas other studies have reported the CVM stage was considered as a continuous variable rather than a
agreement for CVM staging to be substantial.29 Reasons for the poor qualitative variable and that chronological age was evaluated as a
reliability have been attributed to the level of training, clinician expe- continuous linear variable (implying that mandibular growth changes
rience and methods of assessment.29 In the current study, the assess- linearly along with age). In contrast to the present study, Engel et al21
ment of the CVM staging was performed by an expert examiner and concluded that there is no evidence to support the hypothesis that
the intra-observer reproducibility was almost perfect. the CVM method can predict the amount of craniofacial growth in
girls with Class II malocclusion. Our findings do not agree with this
conclusion.
4.1 | Summary of the findings Grey et al22 analysed the CVM method and its relationship to
observed changes in mandibular length during growth. Mixed model
The results of the present study showed that the maturation of the analyses were used to determine the relationships between mandib-
cervical vertebrae, gender and chronological age are related to man- ular length, gender, CVM stage and chronological age. Mandibular
dibular growth in a significant manner. In particular, the greatest an- length represented the response variable, with gender and CVM
nualized increment in mandibular growth was found during the year stage entered as covariates. The interaction between CVM stage
following the appearance of CS 3. The smallest increases were found and gender also was assessed. The results showed that mandibu-
to occur after stages CS 5 and CS 6. lar length was associated with chronological age and to a marginal
As for gender, in general, the increases in mandibular growth in extent with gender. No significant association was found between
male subjects were greater than in females. This difference can be mandibular length and cervical stages. It should be noted, however,
visualized easily in Figure 2 where all the growth curves for males that the differences with the present study were that the outcome
were higher than those for females for all CVM stages. At the age of variable was mandibular length and not mandibular growth incre-
8-9 years, all subjects were either in CVM stage CS 1 or CS 2 and the ments and that chronologic age was evaluated as a continuous linear
increases in mandibular growth were similar in males and females. variable.
From the age of 9 years onward, females already can be in stage CS
3, and they showed greater mandibular increases than males who
are still in CS 1 or CS 2. In general, therefore, females tended to grow 4.3 | Limitations of the study
more than males during the early ages, while males showed longer
and higher growth increases than females during later ages. It should be noted that the lateral cephalograms were collected in
It is interesting to note that the curves CS 1 and CS 2 in females the various Growth Studies from the 1930s to the 1980s. Thus,
and males started from higher values at 8 years, and then, they reached these data could be affected by secular trends.30,31
a minimum around the age of 9 years followed by an acceleration of The final sample was relatively small with respect to the parent
growth. This observation confirms the existence of a prepubertal mini- sample analysed. Moreover, small samples were gathered from di-
mum in mandibular growth that has been described previously.4,5,7 The verse independent growth studies.
greatest increase in mandibular growth occurred at CS 3 for both fe- Another limitation was the lack of validation of the prediction
males and males with annualized values of 3.2-3.3 mm at 12-13 years in model. Unfortunately, it was not possible to validate the prediction
females and with annualized values of 4.1 mm at 13-14 years in males. model on a different sample because all eligible subjects available
The curves for stages CS 4, CS 5 and CS 6 were characterized through the AAOF Craniofacial Growth Legacy Collection Project,
by growth deceleration that is typical of the postpubertal growth the University of Michigan Growth Study and of the Denver Child
phases. Our data also confirmed that mandibular growth diminishes Growth Study were included in this study. No other growth study
substantially earlier in females (around 16-16.5 years) than in males was available.
(around 18-19 years).

4.4 | Clinical implications

4.2 | Comparison with other papers
The results of the current study showed that CS 3 corresponds
The results of the present study differ from those reported by Engel to the stage that precedes the year with the greatest annualized
et al21 and Gray et al. 22 Engel et al21 assessed the performance of increase in total mandibular length in both males and females. Our
6 | FRANCHI et al.

data support the recommendation that CS 3 represents the ideal REFERENCES

stage to begin functional jaw orthopedics in Class II patients, as 1. Perinetti G, Primožič J, Franchi L, Contardo L. Treatment effects of
has been suggested previously by other investigations.1-3 When removable functional appliances in pre-pubertal and pubertal Class
II patients: a systematic review and meta-analysis of controlled
a patient starts treatment at CS 3, after 2 years of functional jaw
studies. PLoS One. 2015;10:e0141198.
orthopedics followed by fixed appliances or aligners, it is likely 2. Perinetti G, Primožič J, Furlani G, Franchi L, Contardo L. Treatment
that he/she will show a postpubertal stage CS 4-CS 5. This aspect effects of fixed functional appliances alone or in combination with
works in favour of a better long-term stability. 3 On the contrary, multibracket appliances: a systematic review and meta-analysis.
Angle Orthod. 2015;85(3):480-492.
if treatment with functional jaw orthopedics is started before pu-
3. Pavoni C, Lombardo EC, Lione R, et al. Treatment timing for func-
berty (CS 1-CS 2), the patient shows a temporary acceleration of tional jaw orthopaedics followed by fixed appliances: a controlled
mandibular growth that is followed by a relapse during the post- long-term study. Eur J Orthod. 2018;40(4):430-436.
treatment period with an overall long-term effect that is similar 4. Björk A. Variations in the growth pattern of the human mandible:
longitudinal radiographic study by the implant method. J Dent Res.
to untreated Class II subjects. 3 Therefore, if the aim is to produce
1963;42(1):400-411.
effective skeletal mandibular changes, the start of treatment with 5. Björk A. Sutural growth of the upper face studied by the implant
removable functional appliances should be postponed until pu- method. Trans Eur Orthod Soc. 1964;40:49-65.
berty. On the other hand, if the correction of the Class II problem 6. Hunter CJ. The correlation of facial growth with body height and
requires mainly dentoalveolar changes, treatment can be initiated skeletal maturation at adolescence. Angle Orthod. 1966;36(1):44-54.
7. Lewis A, Roche AF, Wagner B. Pubertal spurts in cranial base
before puberty. 3
and mandible. Comparisons within individuals. Angle Orthod.
1985;55(1):17-30.
8. Björk A, Helm S. Prediction of the age of maximum pubertal growth
5 | CO N C LU S I O N S in body height. Angle Orthod. 1967;37(2):134-143.
9. Mellion ZJ, Behrents RG, Johnston LE Jr. The pattern of facial skel-
etal growth and its relationship to various common indexes of mat-
The present study developed a prediction model that combined the uration. Am J Orthod Dentofacial Orthop. 2013;143(6):845-854.
information derived from chronologic age, gender and the CVM 10. Fishman LS. Radiographic evaluation of skeletal maturation. A clin-
method to predict mandibular growth. ically oriented method based on hand-wrist films. Angle Orthod.
1982;52(2):88-112.
It was demonstrated that:
11. Flores-Mir C, Nebbe B, Major PW. Use of skeletal maturation based
on hand-wrist radiographic analysis as a predictor of facial growth:
• Cervical stages, age up to the fourth order, and gender, were sig- a systematic review. Angle Orthod. 2004;74(1):118-124.
nificant predictors for the annualized increments in mandibular 12. Hassel B, Farman AG. Skeletal maturation evaluation using cervical
vertebrae. Am J Orthod Dentofacial Orthop. 1995;107(1):58-66.
growth.
13. Baccetti T, Franchi L, McNamara JA Jr. The Cervical Vertebral
• The annualized increment in Co-Gn was significantly greater for Maturation (CVM) method for the assessment of optimal treatment
CS 3 when compared to all other cervical stages. timing in dentofacial orthopedics. Semin Orthod. 2005;11(3):119-129.
14. McNamara JA Jr, Franchi L. The cervical vertebral maturation
method: a user's guide. Angle Orthod. 2018;88(2):133-143.
AC K N OW L E D G E M E N T S
15. O'Reilly M, Yanniello GJ. Mandibular growth changes and matura-
The authors would like to thank Dr Irene Lomonaco and Dr Ginevra tion of cervical vertebrae–a longitudinal cephalometric study. Angle
Batistini for their help in selecting the sample from the AAOF legacy Orthod. 1988;58(2):179-184.
website. 16. Franchi L, Baccetti T, McNamara JA Jr. Mandibular growth as re-
lated to cervical vertebral maturation and body height. Am J Orthod
Dentofacial Orthop. 2000;118(3):335-340.
C O N FL I C T O F I N T E R E S T
17. Cericato GO, Bittencourt MAV, Paranhos LR. Validity of the as-
None of the authors declare any conflicts of interest. sessment method of skeletal maturation by cervical vertebrae: a
systematic review and meta-analysis. Dentomaxillofacial Radiol.
AU T H O R S ’ C O N T R I B U T I O N S 2015;44(4):20140270.
18. Perinetti G, Contardo L, Castaldo A, McNamara JA Jr, Franchi L.
Lorenzo Franchi: conception and design, writing of the manuscript.
Diagnostic reliability of the cervical vertebral maturation method
Michele Nieri: data analysis and interpretation of data. James A. and standing height in the identification of the mandibular growth
McNamara Jr: conception and design, editing of the manuscript. spurt. Angle Orthod. 2016;86(4):599-609.
Veronica Giuntini: conception and design. 19. Hosni S, Burnside G, Watkinson S, Harrison JE. Comparison of stat-
ural height growth velocity at different cervical vertebral matura-
tion stages. Am J Orthod Dentofacial Orthop. 2018;154(4):545-553.
DATA AVA I L A B I L I T Y S TAT E M E N T 20. Szemraj A, Wojtaszek-Słomińska A, Racka-Pilszak B. Is the cervical
Data are available on request from the authors. vertebral maturation (CVM) method effective enough to replace
the hand-wrist maturation (HWM) method in determining skeletal
maturation?-A systematic review. Eur J Radiol. 2018;102:125-128.
ORCID
21. Engel TP, Renkema AM, Katsaros C, Pazera P, Pandis N, Fudalej PS.
Lorenzo Franchi https://orcid.org/0000-0002-2072-460X The cervical vertebrae maturation (CVM) method cannot predict
Michele Nieri https://orcid.org/0000-0001-8770-4622 craniofacial growth in girls with Class II malocclusion. Eur J Orthod.
Veronica Giuntini https://orcid.org/0000-0002-7575-5008 2016;38(1):1-7.
FRANCHI et al. | 7

22. Gray S, Bennani H, Kieser JA, Farella M. Morphometric analysis of on reproducibility of the cervical vertebral maturation method. Am
cervical vertebrae in relation to mandibular growth. Am J Orthod J Orthod Dentofacial Orthop. 2020;157(2):228-239.
Dentofacial Orthop. 2016;149(1):92-98. 30. Aksglaede L, Olsen LW, Sorensen TI, Juul A. Forty years trends in
23. Perinetti G, Contardo L, Primožič J. Diagnostic accuracy of the cervi- timing of pubertal growth spurt in 157,000 Danish school children.
cal vertebral maturation method. Eur J Orthod. 2018;40(4):453-454. PLoS One. 2008;3:e2728.
24. Montasser MA. Craniofacial growth spurt in Class I subjects. Am J 31. Patcas R, Wiedemeier DB, Markic G, Beit P, Keller H. Evidence
Orthod Dentofacial Orthop. 2019;155(4):473-481. of secular trend in mandibular pubertal growth. Eur J Orthod.
25. Morris KM, Fields HW Jr, Beck FM, Kim DG. Diagnostic testing of 2017;39(6):680-685.
cervical vertebral maturation staging: an independent assessment.
Am J Orthod Dentofacial Orthop. 2019;156(5):626-632.
26. Moons KG, Altman DG, Reitsma JB, et al. Transparent Reporting
How to cite this article: Franchi L, Nieri M, McNamara JA Jr,
of a multivariable prediction model for Individual Prognosis or
Giuntini V. Predicting mandibular growth based on CVM
Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med.
2015;162(1):W1-W73. stage and gender and with chronological age as a curvilinear
27. Landis JR, Koch GG. The measurement of observer agreement for variable. Orthod Craniofac Res. 2020;00:1–7. https://doi.
categorical data. Biometrics. 1977;33(1):159-174. org/10.1111/ocr.12457
28. Gabriel DB, Southard KA, Qian F, Marshall SD, Franciscus RG,
Southard TE. Cervical vertebrae maturation method: poor repro-
ducibility. Am J Orthod Dentofacial Orthop. 2009;136(4):478.e1-7.
29. Khajah A, Tadinada A, Allareddy V, Kuo CL, Nanda R, Uribe F.
Influence of type of radiograph and levels of experience and training