llriti.,/t Jourmtl u/' lJrt/w,/olllinll'ul. !.11/IJIJh/IJ./- 10!

Balancing the Books on Orthodontic

Treatment: An integrated Analysis of Change
LYSLE E. JOHNSTON, JR, D.D.S .• M.S., PH.D., F.D.S. ILC.S.(ENG.)
Department ol' Orthodontics and Pediatric Dentistry. Sehoul ol' Dentistry. The University ol' Michigan. Ann Arhur, Michig;m -IMIUIJ-107K.
U.S.A.
Accepted fur puhlicatinn

Abstrad. A metlwtl of ceplw/ometric lllllllysi.\" is pre.H'Illetl in which molar am/ m•t•rjt•t corrt·ctions ttrl' .\'f't'n till' algehmic
.\"IIIII jilcitll skelt•tal wowth tllltl tooth 11/lll't'llll'nt rl'latit·e to lm.ml hom·. The 111111 11/l'ti.\'1/Tt'll/t'l/l
are tlescrihet/, am/ tllllt'llll.\" of.mmmarizing diugram '-is .mggt•stet/.
Index wtm/s: Ccphnlumctric Annlysis. Fncial Growth. Orthodontic Trcntment Effects.

'analysis'. much less a named analysis. In the end. I Jo so

lntrodudion
only for narrative cunvenience. It is perhaps mnre prof-
The cephalometric technique enjoyed more than 60 itably seen as an intuitively appealing. internally logical
Downloaded by [Tufts University] at 07:03 20 July 2016

years of popularity and respect. It has become an enlluring means of characterizing and contrasting treatments
symbol of the orthodontic specialist. This exalted status un their component effects.
has strung theoretical justification: the cephalogram can Although growth and lrt!atment interact in all three
describe the anatomical basis of the malocclusion and planes of space. the analysis is designed specifically to
provide a detailed characterization of the impact of treat- produce a reconciled accounting of <llllern-pustcrinr
ment. In practice. however. there is surprisingly little in change measured at the level nf the occlusion. This direc-
the way of demand for this type of information. When tional emphasis is merely a reflection of the author's
experience teaches that very nearly any treatment is good pt!rsonal interests. rather than a fully informed judge-
enough to permit success in practice. proof of efficacy ment as to the most important plane of space. Others
may seem beside the point. Thus. despite its widely- (sec. fur example. Hans t'l a/., llJlJ4) no doubt will wish to
acknowledged potential. the cephalogram. more often characterize changes in width and height. In any event.
than not. is of limited practil:al significance. much like a the present apprm1ch views the correction of maloc-
chimney sweep's top hat or the shine on a set of study clusion-molar relationship and overjet-as the end result
models. of a series of physical displacements produced hy growth
A by-product of the historical of proof of and tooth movement: displacement of maxill•• relative tn
efficacy is a strange kind of therapeutic immortality: long cranial base. movement of maxillary dentition relative tn
discredited treatments (expansion and bite-_jumping, for maxillary basal bone. translation of mandible rdative to
example) can be resurrected. fitted out with a new theo- cranial base, and movement of mandibular dentition
retical justification. and then sold to the profession as relative to mandibular basal hone.
'cutting-edgt!' alternatives. As noted by Santayana. those More to the point of this communication, the change in
who forget the past arc condemned forever to repeat it. the molar and incisor relationship can be expressed as an
Many dentists. however. art! eager to repeat the past. exact algebraic sum if the component displacements arc
largely because an uncritical acceptance of recycled ther- measured in a comparable manner and each is given a
apies, especially non-extraction recycled therapies. has sign appropriate to its impact: positive if it would tend to
proved to be an effective short-term answer to the 'busy- correct a Class II molar relationship or reduce overjet (as
ness' problem. Others find it difficult to suspend disbelief would he the case. say. with forward growth uf the
and complain instead that 'someone should do some- mandible or mesial movement of the lower molars and
thing'. For tht! more reflective dissenter. the something incisors): negative. if it increases the overjet nr moves the
that someone should do is to gather long-term data. To molar relationship toward Class II (e.g. as with forward
this end. the cephalogmm may at long last be called upon growth of the maxilla or mesial movement of the upper
to fulfill its unrealized promise. both for the good of the dentition). Given this sign convt!ntion. the algebraic sum
patient and fur the integrity of tht! specialty. of the various antero-posterior skeletal and dental effects
The purpose of this communication is to describe a would equal tht! change in molar relationship overjet.
method of cephalometric analysis that has evolved over Accordingly. this accounting can support comparisons of
the past two decades and which is designed specifically to change between treatments and between treatment
measure the dental and skelt!tal effects that combine to phast!s, not only with respect to magnitude. but also
produce the occlusal changes that are the goals of treat- .source (i.e. skeletal or dental). The so-called 'pitchfork·
ment. Because it is based on venerable concepts and diagram (Fig. l) provides a convenient and logical means
because its details surely would be-and probably have of organizing and summarizing the various components
been--duplicated independently by anyone with a serious of change that come together at the occlusal phme.
desire to measure change. I am hesitant to call it an Nott! that apical base change (ARCH). the sum of

(i) I•J% llrilish Orthuolo>nliL" Sno:iL-1)'

 94 L. E. Johnston. Jr. 810 Vo/23 No. 2

Cranial Base
Measured relative to W time
1.
Maxilla (Max)

c.
Upper Incisor
..._..,.._ _ _" (U1)

Skeleton Over jet (111) 2.

(ABCH)
(L1)
Lower Incisor
Lower Molar
3.
Mandible (Mand)
Max + Mand • ABCH
ABCH + U6 + L6 • 6/6
Downloaded by [Tufts University] at 07:03 20 July 2016

R
ABCH + U1 + L 1 • 111
FIG. I "Pitchfork' analysis. The tines of the fork represent skeletal Fru. 2 The elements of superimposition. In row I, it may be seen that
translatory growth relative to cranial base and tooth movement relative to superimposition is an easy and direct one-step procedure when there is no
basal bone. The various skelelal and dental changes sum to produce change change (a to b): however. when there is 'growth' (b to c) one is faced with the
in molar relationship (1\/1\) and overjet (Ill). separate problems (row 2) of registration (the process of 'spindling' the
tracings on some defined point or structure, R) and orientation (the act of
rotating the tracings until some specified angular relationship-usually
coincidence or parallelism-has been achieved). As is illustrated further in
maxillary and mandibular translatory growth relative to row 3. an entirely different pattern of change emerges when different
cranial base, represents the net effect of skeletal growth, registration points are chosen.
usually the amount that the mandible has out-grown the
maxilla. Thus, ABCH plus upper and lower molar move-
ment equals the change in molar relationship; ABCH
plus upper and lower incisor movement equals change in
overjet. In passing, it is important to emphasize that this superimpositions appropriate to the questions at hand.
accounting will always be internally consistent: the com- To this end, it must be emphasized that a given subject's
ponents of growth and tooth movement, no matter how cephalograms cannot be traced casually and indepen-
well or how poorly measured, must add up to the total dently. Rather, they must be traced at a single sitting,
molar and overjet correction. In other words, numbers side-by-side, and in temporally adjacent pairs (time 1 and
from a sloppy analysis will add up just as precisely and time 2; time 2 and time 3; etc.). Each bony detail common
look just as good on paper as will those from a well- to the two films is traced in parallel: a line on one tracing,
executed analysis. Ultimately, therefore, the validity of then the same line, executed in the same way, on the
the data generated by this analysis, indeed, any analysis, second. Given that the analysis is based on the tracing
is proportional to the skill and care with which the re- and superimposition of shared anatomical details that
quisite superimpositions and measurements arc executed. may differ from patient to patient, it may be quite some
A second goal of this paper, therefore, is to present a de- time before commercial cephalometric programs will
tailed description of the technical aspects of the analysis, allow us to rely on digitization to generate the tracings, to
including the author's preferences concerning tracing, perform the superimpositions, and to execute the measure-
superimposition, and the measurement of change. ments. In any event, if the tracings are executed properly,
key radiographic details will match up throughout the
series, and the resulting superimpositions will support an
Superimposition analysis of positional change.
For example, to measure the movement of structure X
The process of measuring skeletal and dental displace- relative to structure Y, one can superimpose-or at least
ment involves, either directly or indirectly, some form of register-on stable bony details in X and then measure
superimposition. Superimposition, in turn, consists of the displacement of Y, or, alternatively, superimpose on
registration and orientation (Fig. 2), both of which must stable structures in Y and measure the displacement of X.
be based on stable reference structures if the changes that lt makes no difference; however, once the change has
we measure are to reflect only bodily displacement and been measured, it usually is expressed as displacement of
not a mixture of displacement and remodelling. A key the structure that craniofacial biologists tell us growing
element of the present technique, therefore, is the execu- more rapidly during the period of observation. In other
tion of co-ordinated tracings that capture stable ana- words, it makes more sense to an orthodontist, if not to
tomical structures in sufficient detail to support the Einstein, to see the maxilla as growing forward relative to
 BJOMa_v IIJIJh

Cranial hem.•
Although S-Na (registered at S or Na) commonly serves
as the basis of an anterior cranial base superimposition. a
number of studies imply that both S and Na arc changed
by local remodelling during the growth period (Ford.
1958; Scott, 195S; Latham, 1972; Mclsen. llJ74). In con-
trast. the literature argues that the hony anatomy from
the anterior half of sella turcica to the region of foramen
caecum and the internal outline of the frontal bonl:' is suf-
Fw. 3 Cranial base superimposition. As dcpich:d here in hc••vicr uutline. ficiently stable to support nwaningful antl:'riur cranial-
only the cranial-base structures between the dashed vcrlicallincs arc base superimposition (De Coster. llJ51; Bji.)rk and
employed for superimpusition. Emphasis is placed un the clements nf de Skieller. Doppelet a/.. llJlJ4 ). Spl:'cilically. Bji)rk and
Coster's "basal line" (de Custer. 1'151 ): the pustcriur half of sclh1 turcica •md Skiellcr ( llJX3) suggest that the following 'natural refer-
structures in the region nr nasion arc ignnrcd. w dcnutcs "wing puinl .. the
point ill which the avcnlgcd t.llltlinc ur the wings the ju!!um. A
ence structures' he employed: the anterior wall of sella
long arbitrary line with crosses on c<lch end---a "fiduciallinc"----is drawn a bow turcica (and its point of intersection with the lower con-
the orbital plates or one tracing. The cnmial-b<ISC superimposition tlwn is used tours of the anterior clinoids). the greater wings of the
In pass this line thruugh In the nthcr tracings in lhc series. sphenoid. the cribriform plate. the orbital roofs. and the
inner surface of the frontal hone (Fig. 3 ).

cranial base. rather than the cranial base moving back-

Downloaded by [Tufts University] at 07:03 20 July 2016

Maxilla
ward relative to maxilla.
Bodily translation of basal hone can m_ove teeth; In case reports. maxillary superimposition is commonly
face changes cannot. The pitchfork analys1s thus reqUires clkcted hv wav of an ANS-PNS orientation in conjunc-
that we measure skeletal change as actual physical dis- tion with- on 1\NS or the lingual palatal
placement. rather than apparent change in the position of curvature. As judged against metallic implants. however.
a landmark due to surface remodelling. Thus. our measure- this method tends to introduce considerable bias. espe-
ments must he executed with respect to landmarks that cially in terms of vertical displacement of the molars and
have a good chance of being physically the same at incisors (Ludcr. llJH I; Baumrind c·t a/.. Nielsl:'n,
both time-points. All too often, this requirement is not 1989; Duppcl c•t a!.• 1994). In contrast. the so-called
met. ·structural method' of B.itlrk (I 955, I t.IM. 1%8) ami Bjt)rk
Change in SNB. for example, commonly is interpreted and Sk icller ( 1972. 1976. JlJ77a.h) seems to provide a use-
as signifying change in position of the mandible relative ful approximation of an implant superimposition. The
to a cranial base superimposition in which we orient on S- structural method in its various forms is based on the
N and registt:r at N. Although SNB is a popular :zygomatic process nf the maxilla. especially its anterior
ment, its validity is compromised by the fact that m surface. Unfortunately. the outline of the anterior surface
growing subjects none of the three landmarks can be of the zygomatic process is not only difficult to sec, hut
assumed to be the samt: through time. The delinitions also too short to achieve reliable control of palatal plane
stay the same. but the points. themselves. probably are angulation (i.e. to provide reliable orientation). Indeed.
physically different. Moreover. given the _re- Nielsen ( 19H9) examined serial ccphalograms from a sam-
modelling on surface landmarks. regional supcnmposttton ple of implanted subjects ami found that the
on stable anatomical details also is required to standard- process was unusable about half of the time. Based on the
ize the position of landmarks in the hones whose dis- bulk of the literature. adequate sagittal control probably
placement is to be measured. can he achieved with best-lit registration on hoth the
For example. if mandibular tracings arc superimposed zygomatic process of the maxilla (right and kft sidL·s
on stable structural details, the position of B (or any averaged) and on the bony anatomical details superior to
other mandibular landmark, real or arbitrary) can he the incisors. The superior and inferior surfaces of the
transferred from one tracing (say. the first) to all others in posterior hard palate assist in orientation. and to mini-
the series, thereby standardizing its position with respect mize the probability of gross errors in antero-posterior
to stable anatomical structures. Given this procedure. registration. care should he taken to ensure that the JYJ'M
change in position of B would rellcct of lissurc of the older tracing lies at or behind that nf the
anterior mandibular basal bone. uncontammatcd by younger (Fig. 4 ).
apparent change due to surface remodelling.
The present analysis of occlusal change employs three
general base. maxilla. Mandible
mandible. These supenmposttton can be executed m
many ways, including those commonly employed in the For the purposes of measuring tooth movement relative
documentation of 'board cases· or the illustration of clini- tu basal hone. mandibular regional superimposition com-
cal reports. Decades of implant research. however. argue monly is effected via a mandibular-plane orientation and
that we can do better. Indeed, properly chosen anatom- a lingual-symphysis registration. Implant studies, how-
ical details can be used to produce superimpositions that ever, suggest that greater validity can he achieved hy
are comparable in many ways to those based on metallic using the mandibular canal, tooth germs (prior tn the
implants. initiation of root formation), and the individual hony
 96 L. E. Johnston. Jr. BJO Vo/23 No. 2
Downloaded by [Tufts University] at 07:03 20 July 2016

F11;. S Mimdihular regional superimposition. Registration is achieved hy

aligning the hony architecture of the facial half of the symphysis; orientation,
hy aligning the mandibular canals or molar Iouth germs. The superimposition
is used to carry both the fiducial line and D point (the centre of the bony
symphysis. hy inspection) through from one tracing to the other.

sitions described here can also be useful in illustrating the

overall pattern of growth. When each tracing has been
executed in appropriate detail, 'fiducial lines' (arbitrary
straight lines several inches long marked on either end
with registration crosses) are drawn adjacent to cranial
base, maxilla, and mandible of one tracing (say, from the
Fw. 4 Maxillary regional superimposition. Registration is hased on the
zygomatic process of the maxilla ('key ridge') and the curvature of the palate middle of the series). The appropriate regional super-
(i.e. structures in the region of R); orientation, on the horizontal structures of imposition then is used to transfer them forward and
the hard palate. Note that care is taken to ensure that the pterygomaxillary backward, pairwise, throughout the series (see Figs 3-5).
fissure of the older tracing is at or he: hind that of the younger. Once again, It is important to note that once a tracing has been
the superimposition is recorded hy an arbitrary fiducial line. Maxillary used in a superimposition, its fiducial lines are inviolate.
advancement relative to cranial ha•c (MAX) is measured at W; Mandibular
displacement relative to maxilla (ABCJ/) is measured at D. Both
For example, after the lines have been transferred from
measurements arc executed parallel to MFOP. tracing 2 to tracing l, the transfer from 2 to 3 may require
that tracing 2 be modified to emphasize anatomical
details common to film 3. However, in this process, its
architecture in the labial aspect of the symphysis (Bjork fiducial lines must never be touched.
and Skicller, 1983: Dibhcts, 1990; Fig. 5). Unfortunately, Once the fiducial lines have been transferred through-
many of these bony details are often absent, difficult to out the series. they serve not only to record the super-
see, or distorted by among-series variation in head place- impositions to facilitate the process of measurement
ment. When, as is sometimes the case, nothing seems to (and, if necessary, re-measurement), but also to provide a
match (except, perhaps, bony details within the symphy- greatly simplified picture of change seen from the van-
sis), the mandibular plane can be used as a substitute tage point of any of the three facial areas (Fig. 6). If, for
orientation line, especially if there has been minimal example, the cranial-base fiducial lines are superimposed,
growth between cephalograms. Indeed, if there has been the separation of the mandibular and maxillary fiducial
little change in overall size, any superimposition that lines represents the translatory growth-both in amount
seems to show marked growth rotation probably is and angulation-of the jaws relative to cranial base.
wrong. In passing, it may be noted that the use of fiducial lines
can be applied to the analysis of other types of change.
The pattern of craniofacial growth For example, one can use regional superimposition with-
in proximal and distal segments to examine the stability
Before the measurement of skeletal and dental change is of mandibular advancement surgery. In effect, because
examined, it should be noted that the three superimpo- each line contains all the information about position of
 llJO M11y /IJI/6 Pitdth•rk t\nulysis '17

analysis of change, in this case, the effects of growth and

treatment
Although the face undergoes widespread change during
orthodontic treatment, only effects that are felt at the
level of the occlusion can have a diret·t impact on the
molar and incisor relationships. The occlusion, therefore,
represents the 'bottom line', the site at which change in
the upper and lower jaws comes together and is inte-
grated. For this reason, the pitchfork analysis measures
change projected onto the plane of occlusion ( i.l'. parallel
i
+ to I he occlusal plane). However, how is the occlusal plane
to he defined? The popular Downs occlusal 'plane'
(Downs, Jl)4H) is drafletl 'hy bisecting the first molar cusp
height anti incisal overbite anti connecting the two with a
straight line'. Unfortunately, the virtue of simplicity is
hough! at the cost of validity: the Downs occlusal linl•'s
angulation is a function of incisor position anti thus is
relatively superficial and unstable. Accortlingly, the pitch-
fork analysis uses a ·functional' occlusal plane ( ,..OP)
B. similar to that suggested hy Jenkins ( 1955) and tie lined as
Downloaded by [Tufts University] at 07:03 20 July 2016

•... the average occlusal plane of the buccal teeth, includ-

ing canine antllirst permanent molar·. As used here, FOI'
is 'a best-lit line passing through the occlusal overlap in
the region of the first molars. premolars. and canines
(especially when premolars have heen extracted). Thus. it
is insensitive to incisor movement, it is of
the hulk of the buccal occlusion, and it is relatively stable
over time, although its angulation with respect to cranial
hase tends to decrease slightly and progressively as the
mandible outgrows the maxilla. Unfortunately, the
functional occlusal plane also tends to he technically
challenging. It can serve as the basis of an analysis of
Ftc;, h The pallcrn of translatory growth (male, non-extraction, ages 11.1-l,
and 33). (A) Sup<·rimpusition in the cranial hasc depicts a mnrc-or-kss typical
change only if great care is exercised during the execution
panern uf maxillarv and mandibular translatory growth. Note that the fiducial of the tracings.
lines (right) ;1 greatly pil'turc of hoth <lispl<tcemcnt and rota- Firstly, it is important to remember that FOP is tlctined
linn. (B) Maxillary sup.:rimpnsitinn prnvid.:s a ditnl pktur.: nf Ih.: mostly hy the premolars and first permanent molars: the
dinenmtial growth of th•• mandihlc (A/lCI/) and th" d.:nroalvcniat cnmpen- second and third molars, and incisors must he ignored.
satinns commonly in lu this path:rn uf grnwlh. Oncl..' agiJin. tlw Secondly, the line (sic) is placed by inspection. either with
fiduciallin"s (right) prnvid.: a d.:ar pictur.· of the fnrwanl rutalion nf
mandible relative In maxilla.
respect to the radio-opacities created hy cuspal overlap
or to the ratliolucencies scattered among cusps along the
line of the occlusion (I prefer to concentrate on the
radiolucencies). ln judging goodness of lit, it is helpful to
slide a strip of acetate inscribed with a single straight line
basal bone available in the tracing, the liducial line between the film and the tracing. The line is positioned
becomes the structure whose superimposition it records. over the huccalocclusion. and, by trial-anu-error manipu-
In the present report. however, the main purpose of lation. used to delineate the occlusal plane. The resulting
fiducial lines is to provide a permanent record of hcst-tit orientation then is copied through to the over-
regional superimpositions used in tlu: measurement ol
lying tracing. In gcnt:ral. the pre-treatmt:nt FOP tends to
skeletal and dental change. cant tlownward a few degrees relative to DOP; at the end
of treatment, however, the two lines commonly coincide.
Measurement of change When a two-film series is to he analvsed the maxillae
are superimposed, anu the two FOP averaged by
In its emphasis on the integration of disparate el.ements: inspection to yield a mean functional occlusal plane
the pitchfork analysis clearl.Y. owes to the tdeas ot. (MFOP). which is passed through to each tracing. This
Wendell Wylie (as cxemphhed by · fhc Assessment ol averaging is aided, once again, hy the usc of a lined
Antero-posterior dysplasia', 1947). Moreover, it borrows acetate strip. When change is to be measured in a series
heavily, as docs Jacobson's 'Wits' analysis (Jacobson, of more than two lllms, the !\'!FOP is ohtaint:d by averag-
197.5), from the analytic methods of D. Harvey Jenkins ing the initial and tina) functional occlusal planes. Once
( 19.55 ), who advocated abstracting maxillo-mandibular the MFOP has been established and transferred to all
relationship as the distance between points A and B tracings in the series. regional superimposition within
measured parallel to the occlusal plane (also sec Popo- cranial hasc, maxilla, and mandible is used to measure
vich and Grainger, llJ5LJ: Harris et a/.. 1%3 ). As will he the various components of anteroposterior change that
seen, this method of measurement also lends itself to an make up the tines of the pitchfork.
 911 L. E. Johnston, Jr. BJO Vo/23 No. 2

Jaw growth/displacement relative to cranial base teeth-in this instance, the first molars and central inci-
sors-constitutes a formidable and, for many, nearly
As used here, regional superimposition is based not on a insuperable technical challenge. The present analysis,
single orientation line and a single registration point, but however, relies heavily on dental measurements. It thus
rather on an ad hoc array of bony details in areas of requires care and skill and beyond
known stability. The present analysis of skeletal change, rary norm of drawing tooth-hke cartoons w1th the a1d of a
however, requires that stable details in both the cranial plastic template and a thick pencil.
base and mandible be reduced to single points whose The measurement of change implies a need to compare
temporal displacement can be measured from the van- two or more cephalograms. Thus, in tracing the outline of
tage point of a maxillary superimposition. . a tooth, one has the option of making a custom template
In the anterior cranial base, 'wing point' (W, the pomt from the best film in the series or, indeed, from parts of
at which the averaged outline of the greater wings of the several films. One can add a long axis, contact points, etc.
sphenoid crosses jugum sphenoida/e, Knott, 1969; see also and then use best-fit superimposition on the film to trans-
CBR, Wieslander, 1963; SE, Riolo eta/., 1974) commonly fer the template outline to each tracing in the series.
serves as a registration point and is used here as the Although this method does not guarantee that the teeth
cranial-base reference point from which a.nd will be rendered correctly, it standardizes details that
mandibular displacement are measured .. To should remain constant (tooth form, size, long-axis orien-
both its validity and reliability, the antenor-cramal-base tation, etc.) and thus serves to optimize the measurement
fiducial line (i.e. the full array of natural reference struc- of change within a given series.
tures) is used to pass W from one tracing (for Tooth movement is measured parallel to MFOP from
nience, the middle tracing) through to the others m the the vantage point of a superimposition within either the
Downloaded by [Tufts University] at 07:03 20 July 2016

series. maxilla (Fig. 7) or mandible (Fig. 8). Molar crown

To measure displacement of the maxilla relative to movement is measured at the mesial contact point: root
nial base (MAX), the maxillary fiducial li_nes supenm- movement (if separate estimates of bodily and tipping
posed and the separation of the W pomts ts measured components are desired) from the point at which the long
parallel to MFOP (Figs 3 and 4)._ can be mea- axis is crossed by a line drawn between the apices of the
sured from a cranial base supenmposthon; however, as buccal roots. Crown movement is the algebraic sum of
will be seen, a maxillary superimposition is used here root movement and a component due to tipping. Thus,
because it can generate all three measurements of skele- the tipping component can be estimated by subtracting
tal growth (vide infra). In passing, it should be noted root movement from crown movement. The same pro-
maxillary rotation impacts measurements of both cedure can be applied to the incisors; however, it may be
Iary and mandibular 'clockwtse less important to resolve separate tipping and bodily
rotation increases both; anttclockwtse rotat1on effects here than in the molar region. For most purposes,
both. On average, however, maxillary rotat1on IS it probably is sufficient to measure the displacement of
minimal and thus its effect on the analysts 1s probably the incisal edges of the upper and lower central incisors.
negligible. . . Because the present analysis deals only with antero-
Next, the separation of D-points _m th_e posterior change, Bjork's superimposition procedure can
two tracings by mandibular regional supenmposttton) 1s be modified to eliminate the vertical component of
measured parallel to MFOP (Fig. 4). This measureme!lt mandibular tooth movement, thereby simplifying the
represents apical base change the .growth/dts- measurement of antero-posterior change.
placement of the mandible relat1ve to maxtllary basal As has been described, stable structural details-sym-
bone. Given the usual pattern of facial growth in whic_h
the mandible ·outgrows' the maxilla, ABC:fl
positive. It should be emphasized in passmg that dts- Maxilla
placement of D-point can he due to growth, a functtonal
shift, or, more probably, some mixture of the two. Unfor-
tunately, the cephalometric technique offers no reliable
means of differentiating between the two types of
change. . .
Whatever its sign (and cause), mandtbular Molar
ment relative to cranial base (MAND) then can be esti- movement
mated by subtraction: MAND = ABCII - MAX. For
example, if the maxilla underwent 3 mm of forward trans-
latory growth, and the mandible out-grew it by 4 mm,
MAND = 4 - (-3) = 7 mm. These three
abstract the sagittal growth of the jaws and, together wtth
tooth movement, account for change in molar relation-
ship and overjet. Fl G. 7 Maxillary tooth mnvemcnt, measured parallel to mean functional
occlusal plane (MfOI') with fiducial lines superimposed. Note that total molar
movement (actually, the displacement of. the mesial contact point) can be
Tooth movement relative to basal bone decomposed into a component due tn root movement (measured point
on the lung uxis at the level or the upiccs) and a cnmpnnent due tu llppmg,
To a specialty that no seems take which can be calculated by subtraction: tutal- budily = tipping. The same
metrics seriously, the executton of tracmgs that mclude approach can he applied to the muvcmentur I he incisors.
 BJO M11.v 1996 l'ilchfnrk A nalysb •JIJ

Molar
movement

MFOP

Overjet
Correction
FICi. H Mandihular tooth movement, mcusurcd frnm u D-pninl-pcrpcndicular
registration and a ori.:lllatinn. As with the
maxillary molars. mandihular first-molar mo\'emcnl c>m he seen as
Downloaded by [Tufts University] at 07:03 20 July 2016

a comhinatiun of tipping and hudily movement. Note thai this sclwmc nf

measurement dues not employ the mandihuhor liduciallinc,. /\It hough tho"
appruach simplifies the measurement of unlcru-pnslerior change, it greatly
underestimates vertical change. Lower incisor movement is measured from
the same superimposition (sec Fig. 7). 1:1« i. 'J ( >vcrjcl,.:hanp.L'. lhl' ••f' th\:'
Ieeth in nne j''" (hcrL", the :nullht'll nricniL"ll .\f/·"0/'. ( 'h:III_I!L'
in o\'eqcl "measured parallel to Ml-"0/'as thL· separation of llw inds:ol cllr:cs
physis, mandibular canal, molar tooth-germs-arc used uf the tcclh in the nppu!'!.ik jaw. Thi!'l appru:u:h :dsu i"' U!'IL'll lu cslinmh:
to establish a mandibular superimposition. Although this in molar rehninnship hy mcaMJrinp, lhc !'\L·p••ralinn pninl!\
nf the muhm.. in jaw aflcr un1hc nh.·sial points of thL'
superimposition is useful in examining the overall pattern
first molars in the nlher.
of facial growth, it provides a technically awkward basis
for the measurement of mandibular tooth movement.
Accordingly, the mandibular superimposition is used to
pass a single D-point throughout the series. Tooth move- point of the other; the change in overjet, hy registering nn
ment then is measureJ with reference to an orientation the averaged incisal edge of the upper or lower incisors
on MFOP and a registration on a perpendicular from and then measuring the displacement of the averaged
MFOP erected through D (Fig. R). Displacement l>f the incisal edge in the other arch (Fig. 9). If the individual
crown (mesial contact point) and roots of the averaged components of growth and tooth movement do nut add
first molars and the incisal edge of the averaged central up to the direct measurements of change executed at the
incisors is measured parallel to MFOP. As may he seen, occlusal plane (say, to within 0·2-0·3 mm), the measure-
orientation along M FOP requires a rotati(lfl of the corpus ments arc re-done. Given this requirement of internal
that facilitates the measurement of molar movement. hut consistency, it is helpful In execute the various measure-
which eliminates the buccal-segment eruption that com- ments with the aid of a magnifying glass and l'lcctrnnic
monly occurs secondary to forward growth rotation of vernier calipers.
the mandible. Given the present emphasis on antero-
posterior change. the resulting ease of analysis more than
offsets the loss of vertical information. Reliahility
The reliability of this technique as a whole can he esti-
Molar relationship and overjet c:hangt' mated from double-determinations. In past stm.lics of
Class II treatment, 5-10 per cent of the subjects were
As has been noted earlier, the present analysis, in common selected at random, and their cephalograms retraced and
with many other cephalometric analyses, is easy to do, re-measurcd. Based on Dahlberg's formula (Dahlberg.
but hard to do well. Skill and care are important safe- IY40; Houston, IIJH3), Slh == Jr.tY!2N, where D is the
guards; however, it is important that an additional step he difference between douhlc determinations, pooled esti-
taken to avoid major errors. Specifically, the change in mates of error standard deviations for the various
overjet and the molar relationship obtained from of skeletal and dental change arc summarized
summing the individual tines of the pitchfork should he m Ftg. I 0. As may he seen, most standard deviations lie in
compared with direct measurements obtained from an !he of 0·5-0·7 mm, a level of reliability that derives
MFOP superimposition. lrom a system of checks and balances designed to
The change in molar relationship is measured hy regrs- mmumze gross errors.
tering on the mesial contact point of one molar (upper or All analyses-including those that turned uut to he
lower) and then measuring the separation nf the contact part of the error study---were performed hy a learn. One
 HKl L. E. Johnston, Jr. 810 Vo/23 No. 2

oral traction? How does the effect of a given appliance

0.64 differ in children and adults? Does premolar extraction
'dish in' the profile? If so, how much? What is the
apparent impact of normal growth (i.e. of excess
mandibular growth) on tooth movement during and after
treatment?
A preliminary answer to these questions can be
inferred from Cohlmia ( 1978), Johnston (1986), Johnston
eta/. (1988), Luecke and Johnston (1992), Paquette eta/.
0.90 0.53 (1992), Luppanapornlarp and Johnston (1993), and
Livieratos and Johnston (1995). Many clinicians, how-
0.67 ever, see averages as largely irrelevant to clinical practice.
They argue that means are for academics and that any-
thing short of an estimate of the individual impact of a
given treatment is a gross over-simplification. This exact-
ing criterion requires that clinical comparisons reflect not
1.09 only the average impact of treatment, but also the type of
patient and the intentions and skill of the orthodontist.
N • 73 Unfortunately, it is not obvious how this complex inter-
action is to be modelled. Indeed, the sorry state of the
after Dahlberg, 1940
Downloaded by [Tufts University] at 07:03 20 July 2016

growth-prediction literature implies that so complete an

fl 0 . 1o Reliability. Error standard deviations calculated from 73 sets
analysis may never be possible. Thus, the ultimate signifi-
double determinations derived from randomly-sckc!cd two- and three-him cance of the individual-variation argument is that it tends
series. to provide a ready excuse for ignoring the literature. In
the end, however, the argument that mean effects are
irrelevant ('I treat my patients one at a time') is some-
person executed the tracings and establish7d the thing of a 'red herring'.
lines, then a second worker checked t.he tracmgs, For example, to a learned specialty that cannot decide
not only for the rendering of detail and land- whether functional appliances are a substitute for surgery
mark localization (condyles, ponon,. orb1tale, or, instead, just another means of moving teeth, valid
but also for the details of the regwnal supenmposJt!On estimates of central tendency would be more than ade-
(registration and orientation). The second worker was quate to the task of settling the argument. If the average
blinded as to the nature of each series and thus did not monthly effect of a functional appliance is 0·1 mm of
know when error-study tracings were being checked. is extra mandibular advancement (see, for example, Johnston,
assumed that this collaborative approach to 1986), the fact that some patients do a bit better and
would, at the very least, serve to detect gross outliers. some do a bit worse is beside the point. Surgery features
Accordingly, the errors reported here probably are effects that are nearly an order of magnitude greater than
what smaller than would be achieved by an mvestlgator can be achieved with functional appliances. In terms of
working alone (see, for example, Keeling eta/., 1993). In treatment planning, therefore, the two treatments appar-
any event, power calculations based on the present data ently arc not interchangeable.
imply that samples of about 30 have a reasonable proba- I would argue, therefore, that many of the most vexing
bility (say, 80-90 per cent) of resolving treatment effects questions in orthodontics can be answered, at least to a
in the range ofO·S-1·0 mm. first approximation, by the kind of data that the present
analysis generates. It remains to be seen whether or not
there is a place for data-driven hypothesis testing in
Synthesis today's orthodontic marketplace.
This paper has presented a method of cephalo'!letric
analysis designed to provide an integrated accountmg of
the source-skeletal growth or tooth Acknowledgements
antero-posterior change. Each of the components 1s Supported by N.l.D.R. grant DE08716, and by_
measured separately; however, as a group, sum _to from the University of Michigan Orthodontic Alumm
provide a complete 'explanation' of changes m overjet association and from other friends of the Department.
and molar relationship. This analysis, therefore, has the The author wishes to thank Drs James A. McNamara, Jr,
potential to characterize the (i.e. average) and Carroll-Ann Trotman for their helpful suggestions
of a variety of appliances and tactics. In the process, 1t during the preparation of this manuscript.
should provide data of direct significance to a number of
dilatory controversies that should have been settled long
ago. References
Do treatments with functional appliances feature more Baumrind, S., Korn, E. L., Ben-Basli8t, Y. and West, E. E. (1987a)
skeletal change than do conventional fixed-appliance Quantitation of maxillary remodeling. I. A description of osseous
treatments? Does anchorage preparation preserve anchor- changes relative to superimposition on metallic i!"plants, .
age? Do 'straight wire' appliances '?urn' anchorage? American Joumal of Orthodontics and Dentofacwl OrthopediCS, 91,
What are the penalties incurred by a fmlure to use extra- 29-41.
 HJO M11y /Wfl

Baumrind, S., Kom, E. L., Ben-Bassat, \',and West, E. E. llans, M. •• Kishiyama, C., Parker, S. 1-1., Wolr. Ci. H. and Noach-
(1987b) tar, H. ( 1994)
Quantitation of maxillary remodeling. 2. Masking of remodeling ( 'cplmlnmetric evuluation nf two treatment l'ur deep over-
effects when an ·anlllomical" method uf superimposition is used in bite correction,
the ahscnce of metallic imphmts. rlnglt· Ortlrmlotrmr, 64, 205- 74.
Amc•,-it'fllt Journal of Orthm/muics 111111/Jc•mofadu/ Ortlwpt•,/in. 91, llarris. J. t:., Johnston, L E. and Moyers, K. t:. C1%..\1
463-474. A cephalumetric template: its cnnstrul·tion :md clinicul
Bjiirk, A. (1955) Anwrictm lou malo{ ( hthllllontin, 49, 24') ·.f1.l
Facial growth in man. studied with the aid of meH1IIic implants. Uouston, W. J. II. ( 19K31
AL·ta Odmuolol{im Sctmdimll'im. 13, lJ-34. The analvsis of errors in orthodontic mcasurcnll'nls.
Bjork, A. (1964) Aml'fil'll;t lou mal of ( Jnhotlomici, K.\, .lX2 ·90.
Sutural growth of the upper face. studied hy the implant Jaacnbson, A. ( 197!il
method. The ·Wits' appraisalnl' juw disharmony.
Eumpt•tm Ortltotlmuil· Socit·ty Tnm.mctiom, 4CI, Amt•rinm Jmmwl o{ t Jrrlrotlomin, 67, 12:'i I.'X.
Bjiirk, A. (1968) Jenkins, U. II. C19!i!i)
The usc of metallic implants in the study uf facial in children: Analysis ol' urthodumic ddormity latcrall'L'Jlhalumetril·
method and :1pplieation.
Amt•rimn Joumul of (n ..,·.J. 29, Anu·rinm Jmmwl o{ t Jrtlwtlomin, 41, 44:?
Bjork, A. and Skieller, V. (1972) Johnston, 1.. t:., Jr ( 19KII)
Facial development and tnuth eruption. An implant study at the 1\ comparative analysis of II lrcalnll'nls.
of puhcrty. In: Vi!,!.. P. S. and Rihhl'lls, K. /\. (L•ds). Sril'lll'l' ,,,/ ( 'litrinrl
Anll'rimn Joumulof Ortltmlmuic.l, 112, J3lJ-3X3. Jtulgllll'lll in (
Downloaded by [Tufts University] at 07:03 20 July 2016

Bjiirk, A. and Skieller, V. ( 197f) Ann Arhur, C'cnll"r l'or lluman ( illlWih amll>evclupnwnt. pp.
Postnatal growth :md dcvelopml·nt of the maxillary eumplex .. IIH--1-IX.
In: .1. A. McNamam. Jr (cd.). /·iu·/tln tl.ffecting the ( •'mll'th of rh•· Johnston, 1.. Jo:., Jr. Un, S-S.and l'eng, S. I IIJKKI
Mitl/irl't', No. n. Craniofaci:1l (irnwth Series. (o,s: a comparative analysis.
Center fur Human (iruwth and l>evelopm.:nt. The llniwrsil\'llf Jmmwl of rlt•· "fll"<'t'•/ Fomultlllll/l, 16, 2.l :?7.
Michigan, Ann 1\rhnr. pp. nl-lJlJ. 1\eelinJl, S. U., Caha!>sa, S. H. and KinJl. (j, J. ( 199.\1
Bjiirk, A. and Skieller, V. (1977a) and random errors assnciatl·d with Johnston·, Cl'phalo-
analysis of th.: mandible. nictric analvsio;.
of thl' Eumpt'tlll Ortluulonlil' Socit·t.\', !i3, /lritMt Jou;lllil o{Ortlrotlomin. 211, 101 107.
Bjork, A. and Skieller, V. (1977hl Knoll, V. II. ( 19691
Growth of the maxilla in three dimensinns as rev.: a led radio- Omngcnetk change nf l"nur cranial hasc in
graphically hy th.: implant mcthnd. (imll'rlr. 33, 123--t2.
llriri.1"/t lmmw/ of 4, 53··h4. Knnu,v.B.(l971l
Bjiirk, A. and Skieller, V. (19KJ) in cranial hasc or human mak•s ami from
Nunuill and growth nf the mandihle. A synthesis of age ti YL'IIrs tn early adulthood,
longitudinal ccphalomctrio.; implant ov.:r a p.:riod of 25 years. £iroll'tlr. 35, 145-SK
l:'um/ll't/11 lou mal of !i, 1-46. J.atham, K. A. ( 1972)
Cohlmia, (j, T. ( 197KI The sella point and Ihe postnatal nr the hum:m cranial ha'L',
The relative cuntrihutinns of growth and treatment in Amerim11 Jmmwl of OrtlrotltmticJ, 41, I h2.
nrthudnntically l'llrrech:d ('lass II malucclusiun. . . . Livieratos, t', A. and Johnston, 1.. r•:., Jr ( 199!i)
M.S. D. Thesis. Department of Orthodontics. Saint Lums Umverstly. A cnmparbon nr nne- and nun-extraction in
St. Louis. Missuuri. nmtched Class II samples.
Amt•rimn Joumul of 11111/ Dl'/lto./itdul ( Jrtltof't'tlin, IIIII,
Uahlberg, (194111 IIX.I.H.
Stati.wiml Metlttulx for Mnliml awl /Jio/ogiml Strulnrts,
l.uder,ll. V. (IIJHI)
lntcrsciencc Puhlic:.tiuns. New York.
F.ITccls of activator II"L'almcnl fur lhL' nccurrL·ncl" nf two
l)e Coster, L. (19511 or rL"lll:lilln,
llcreditary potentiality versus mnhicnt l'aclnrs. .. . Eumpt•tm Jmmwl of ( Jrrhmlomic.,, 3, 2115- 2.?2.
ErtmJWtlll Ortlrodomic Socit'l.\', Rt'porr of the Twemy-Fi.ftlr { ongrt•.,.,,.
l.uecke, P. K, Ill, and Johnston, 1.. t:., Jr ( IIJIJ2)
Repnrlnf the Twcllly-Sevcnth ( ::!27-2.,4.
The effector maxillary lirM-hil'Uspid cxlractiun and incisur
l)ibhets, J. 1\1. II. (IIJIJO) relwetion on mandihulm· position: testing lhl' celllral uf
A method for structumlmandihular superimpnsitinning, · runL'I ion a lnrl hndunt ·.
Anll'rimn Joumal of mtd llt·llrojircilrl ( 97, Amerinm Jmmwl of Onlrotlomin· tllltll>l'luojitl'itrl Orthopt•tlit'I, lUI,
M-7J. 4--12.
Doppel, U. M., Damon, W. M.. Joondeph, (),H. and Little, H. M. Luppanapnrnlarp, S. and Johnslon, 1.. F.., Jr ( 1993)
(1994) The effects of prcmnlar-cxtraction trc:llllll'lll: a lun!,!.·lerm cnmpm·i-
An investigation of maxill<try superimposition techniques using son of outcomes in 'clear-cur exlraction and nonextractinn Class II
metallic implullls. . . . . patients.
Americtm Joumal of OrrlrmlmuicJ am/ l>t•mo.ftrcwl Ortlwpetltc.l', IO!i, Angle Ortlwdomi.\'1, 63,
161-lh!!. Melsen, II. ( 19741
Uowns, W. B. (ICJ411) The cranial hasc.
Vuri:lliuns in facial relationships: their signilkance in trealmcnt and Acta Otlontologinr Snmtlirrm·inr, 32 (Suplllellll'nl h2). 1-126.
prognosis, Nielsen, I. L. (1989)
Amt•rican Journal Orrltodtmrics, 34, Rl2-40.
Maxillary superimposition: a eom1larisnn of three nwthods fur
Ford, t:. 1-1. H. I 19!iKI cephalometric l"valuatinnnf growth ami treatment
Growth uf the human cranial hasc, ilml'rinm lmmwlof umllklllo/irl'iul ( Jrtlu•i•••tlic.,·, 95,
Amt•rinm Jormml of ( 44, 4:?2-4.11.
 102 L. E. Johnston, Jr. 810 Vol 23 No. 2

Paquette, D. E., Beattie, J. R. and Johnston, L. E., Jr (1992) Scott, J. H. (1958)

A long-term comparison of non-extraction and bicuspid-extraction The cranial base,
edgewise therapy in 'borderline' Class II patients, American Journal of Physical Anthropology (n.s.), 16, 319-348.
American Journal of Orthodontics and Den to facial Orthopedics, 102, Wieslander, L. (1963)
1-14. The effect of orthodontic treatment on the concurrent development
Popovich, R. and Grainger, R. M. (1959) of the craniofacial complex,
One community's orthodontic problems, American Journal of Orthodolllics, 49, 15-27.
In: R.E. Moyers and P. Jay (eds.), Orthodontics in Mid-Century, Wylie, W. L. (1947)
C.V. Mosby, StLouis, pp. 192-246. The assessment of anteroposterior dysplasia,
Riolo, M. L., Moyers, R. E., McNamara, J. A., Jr, and Hunter, W. S. Angle Orthodontist, 17,97-109.
(1974)
An atlas of craniofacial growth,
Center for Human Growth and Development, Ann Arbor.
Downloaded by [Tufts University] at 07:03 20 July 2016