Objective Diagnosis of Micrognathia in the
Fetus: The Jaw Index
DARIO PALADINI, MD, TIZIANA MORRA, MD, ADELE TEODORO, MD,
AGATA LAMBERTI, MD, FORTUNATO TREMOLATERRA, MD, AND
PASQUALE MARTINELLI, MD
Objective: To provide an objective and accurate tool to
diagnose micrognathia in the fetus.
Methods: The anteroposterior and laterolateral diameter of
the mandible were measured in 262 normal fetuses between
12 and 37 weeks’ gestation and plotted against gestational
age and biparietal diameter (BPD). The jaw index (antero-
posterior mandibular diameter/BPD 3 100) was then tested
against the usual subjective method for diagnosing micro-
gnathia, consisting of the evaluation of the facial profile, in
a population of 198 malformed fetuses, 11 of which had
micrognathia at necropsy or birth.
Results: The mandibular growth was linearly correlated
with gestational age and BPD. Using a cutoff level of less
than 23, the jaw index had a 100% sensitivity and 98.1%
specificity in diagnosing micrognathia, in comparison with
72.7% and 99.2% shown by the subjective evaluation of the
fetal profile. With a cutoff of 21, it yielded a positive
predictive value of 100%.
Conclusion: We demonstrated the linear relationship be-
tween mandibular growth and gestational age or BPD. In
addition, we validated the jaw index as an objective tool for
diagnosis of micrognathia in the fetus. (Obstet Gynecol
1999;93:382– 6. © 1999 by The American College of Obste-
tricians and Gynecologists.)
Facial congenital anomalies are frequently associated
with chromosomal aberrations and genetic syn-
dromes.1,2 In this group of malformations, there are
many anomalies the diagnosis of which relies almost
entirely on the subjective evaluation of the facial profile;
micrognathia is one of these, although recently the
growth of the fetal mandible throughout pregnancy has
been evaluated and normative data have been provid-
ed.3,4 Micrognathia is associated with chromosomal
abnormalities in 66% of cases,5 and a variable degree of
micrognathia has been reported in over 80% of cases of
From the Departments of Obstetrics and Gynecology and Pathology,
University Federico II of Naples, Naples, Italy.
382 0029-7844/99/$20.00
PII S0029-7844(98)00414-1
trisomy 18 and triploidy at autopsy.1,2 Several attempts
have been made to identify objective criteria to diag-
nose micrognathia after birth by x-ray or computed
tomographic scan.6,7
The aims of this study were to construct a growth
chart of the mandible throughout pregnancy, to identify
an objective criterion by which to diagnose microgna-
thia in utero, and to validate it in 11 cases of microgna-
thia detected in utero and confirmed at necropsy or
after birth.
Materials and Methods
All pregnant women between 12 and 37 weeks’ gesta-
tion consecutively referred to our institution from Jan-
uary 1996 through March 1996 for routine second- and
third-trimester scan were evaluated for possible enroll-
ment in the study group. The design of the study was
cross-sectional and therefore each patient was scanned
only once. Entry criteria included singleton pregnancy;
gestational age confirmed by an early second-trimester
biparietal diameter (BPD) measurement; fetal abdomi-
nal circumference greater than the 10th percentile for
gestational age, to exclude the presence of fetal growth
restriction (FGR); absence of congenital anomalies; nor-
mal uteroplacental Doppler velocimetry, to exclude
patients at risk of developing FGR; and the absence of
maternal diseases possibly affecting fetal growth (dia-
betes, pregnancy-induced hypertension, and other sys-
temic diseases). Of the initial 314 women, 52 (16.6%)
were excluded, leaving 262 cases in the study group.
Exclusion criteria included persistent unfavorable fetal
lie in 14 (4.5%), pregnancy dating discordant with last
menstrual period in 10 (3.2%), fetal malformations in
nine (2.9%), abnormal uteroplacental Doppler velocim-
etry in five (1.6%), maternal diseases in five (1.6%), FGR
in five (1.6%), and twinning in four (1.3%). Ultrasound
Obstetrics & Gynecology

Figure 1. Transverse scan of the fetal mandible. A to A 5 laterolateral
diameter; B to C 5 anteroposterior diameter.
examinations were performed by two authors (DP and
TM) using TOSHIBA 270A and TOSHIBA Powervision
ultrasound systems (TOSHIBA Corp., Tochigi-Ken, Ja-
pan) with 3.75 MHz and 3.75– 6.0 MHz multifrequency
convex probes, respectively.
The size of the fetal mandible was assessed on an
axial plane at the base of the cranium just caudad to the
lower dental arch, where the whole horseshoe mandible
is imaged. Anteroposterior and laterolateral diameters
were measured as follows: the laterolateral diameter
was traced joining the bases of the two rami; the
anteroposterior diameter from the symphysis mentis to
the middle of the laterolateral diameter (Figure 1). Care
was taken to achieve the correct plane and to avoid
inadvertent partial inclusion of the rami within the
calipers. Mandibular diameters were plotted against
gestational age and BPD as independent variables to
build the growth charts. The jaw index was then calcu-
lated as follows: anteroposterior mandibular
diameter/BPD 3 100. In the first 20 fetuses, intraob-
server and interobserver variability were assessed by
calculating the Cronbach alpha reliability coefficient for
two measurements taken from the same operator (DP)
and from the two operators (DP and TM) on different
frames. In particular, the first operator (DP) measured
the mandibular diameters as described above and noted
the values calculated on two different frames (intraob-
server variability); then, the second operator (TM) re-
peated the scan and remeasured the diameters on a
newly acquired frame, not knowing the results of the
measurements from the former operator. The interob-
server variability was assessed by comparing the first
set of measurements by the first operator with the set of
measurements taken by the second operator. Twelve
fetuses were scanned by DP first, eight by TM first.
VOL. 93, NO. 3, MARCH 1999
Regardless of who scanned the patient first, intraob-
server variability was evaluated on repeated measure-
ments taken by DP only. The Cronbach alpha reliability
coefficients were 0.97 for intraobserver variability and
0.90 for interobserver variability, indicating a good
repeatability of the measurement. Statistical analysis
was performed with the statistical package SPSS 7.5 for
Windows 95 (SPSS Inc, Chicago, IL). The data gathered
from the 262 normal fetuses were submitted to regres-
sion analysis using gestational age and BPD as indepen-
dent variables. The best-fit curve (among linear, qua-
dratic, cubic, and exponential) was selected to express
the relationship between mandibular diameters and
BPD or gestational age. Confidence limits were calcu-
lated for all variables. Percentiles were developed for
the jaw index. The Mann-Whitney U test was used to
compare jaw index values between normal fetuses and
fetuses with micrognathia. P , .05 was considered
statistically significant.
Once the reference curves and the jaw index had been
developed, they were prospectively tested against a
population of 210 fetuses referred to our unit between
April 1996 and December 1997 for confirmation and
treatment of congenital anomalies. No confirmation of
the prenatal diagnosis could be gathered in 12 cases,
which were excluded from the study, leaving 198 cases
for analysis, 11 of which had micrognathia. In these
fetuses, the diagnosis of micrognathia was first subjec-
tively hypothesized on the basis of the evaluation of the
facial profile and then validated using mandibular
growth charts and the jaw index. On the basis of the
data from the normal fetuses, a cutoff value of 24,
representing the fifth percentile, was prospectively cho-
sen to test the diagnostic accuracy of this variable.
Autopsy reports and pediatric charts were used for
confirmation. Armed Force Institute of Pathology crite-
ria were used to diagnose micrognathia in postmortem
examinations.8 Maternal mean age 6 standard devia-
tion was 29 6 5 years in the group of 262 normal fetuses
and 31 6 0.6 years in the group of malformed fetuses.
Mean parity was 0.8 6 0.8 and 0.9 6 0.9, respectively.
All women were white.
Results
The best equation describing the relationships between
mandibular diameters and gestational age or BPD in the
population of 262 normal fetuses was the linear one
and, for both diameters, the correlation with BPD was
stronger than the correlation with gestational age, as
evident from the lower regression B coefficients in Table
1. The reference curves for mandibular diameters
throughout pregnancy, obtained from the same popu-
lation, are shown in Figures 2 and 3. On these curves,
Paladini et al Micrognathia in Fetal Life 383
Table 1. Linear Regression Analysis of Mandibular
Diameter Compared With Biparietal Diameter and
Gestational Age as Independent Variables
Dependent Independent Confidence
variable variable R2 B* (SE) limits

AP diameter BPD .822 0.33 (.10) 0.31, 0.35

AP diameter GA .806 0.95 (.29) 0.89, 1.01
LL diameter BPD .935 0.47 (.11) 0.45, 0.49
LL diameter GA .868 1.37 (.33) 1.30, 1.43
SE 5 standard error; AP 5 anteroposterior; BPD 5 biparietal
diameter; GA 5 gestational age; LL 5 laterolateral.
* Regression B coefficient. P , .001 for all regressions.

mandibular measurements of the fetuses with con-

firmed micrognathia are overlayed. The data for these
11 fetuses with micrognathia, including gestational age
at diagnosis, associated malformations, karyotype, and
pregnancy outcome, are given in Table 2. As evident
from Figures 2 and 3, micrognathia affects the growth in

Figure 3. Growth chart for the mandibular laterolateral diameter

(mean and 95% confidence limits) plotted against gestational age (wk)
and biparietal diameter (mm) (crosses 5 normal fetuses; circles 5
fetuses with micrognathia).

the sagittal plane more than that in the coronal plane.

Figure 2. Growth chart for the mandibular anteroposterior diameter
(mean and 95% confidence limits) plotted against gestational age (wk)
and biparietal diameter (mm) (crosses 5 normal fetuses; circles 5
fetuses with micrognathia).
Data regarding the jaw index are shown in Figure 4 and
Table 3. This index allows a clear separation between
micrognathic and normal fetuses. Only one fetus with
micrognathia had a jaw index greater than 21, and only
two fetuses without micrognathia had a jaw index less
than 22. The diagnostic accuracy of this objective
method with different cutoff levels is compared with
subjective evaluation of the fetal facial profile in Table 4.
Using the latter method, two false-positive and three
false-negative diagnoses of micrognathia were re-
corded. Among the false-negative diagnoses, one case is
included in which the fetal profile could not be evalu-
ated because of the presence of both an extremely
unfavorable fetal lie and oligohydramnios. In this case,
the diagnosis of micrognathia was based only on the
measurement of the jaw index and was confirmed after
birth.

384 Paladini et al Micrognathia in Fetal Life Obstetrics & Gynecology

Table 2. Characteristics of the 11 Fetuses With Micrognathia Table 3. The Jaw Index: Statistics and Percentiles in the
Confirmed at Necropsy or After Birth Normal Population of 262 Fetuses
GA at Statistic Value
Case diagnosis
Mean (confidence interval) 32.2 (31.6, 32.8)
no. (wk) Associated malformations Karyotype Outcome
Standard deviation 4.9
1 18 VSD, PKK, arthrogriposis Trisomy 18 TOP Median 32.3
2 20 VSD 1 aortic arch Trisomy 18 TOP Range 21–51
interruption, bilateral Percentiles
radial aplasia 5th 24.0
3 21 VSD, bilateral clubfoot, DEL 4p TOP 10th 26.0
FGR 25th 29.4
4 21 Hydrocephalus, NTD, 46, XY TOP 50th 32.3
bilateral clubfoot 75th 34.8
5 21 AVSD 1 coarctation, Trisomy 18 TOP 90th 37.9
oesophageal atresia, 95th 41.0
arthrogriposis
6 21 Campomelic dysplasia 46, XY TOP
7 21 VSD, bilateral radial Trisomy 18 TOP
aplasia skeletal dysplasias, and primary mandibular disorders.
8 22 VSD, exomphalos, Trisomy 18 TOP The last group of diseases includes the Robin anomalad,
bilateral CPC which also has been diagnosed prenatally,9 the domi-
9 23 VSD, Dandy-Walker, Trisomy 18 TOP
bilateral clubfoot
nant and recessive forms of the mandibulofacial dysos-
10 35 Coarctation, ambiguous Trisomy NND tosis, or Treacher-Collins syndrome,10,11 and mandibu-
genitalia, clinodactily, 16p loacral dysplasia.12 Among skeletal dysplasias, a
FGR variable degree of mandibular hypoplasia can be
11 36 VSD, cerebellar Trisomy 18 NND present in campomelic dysplasia13 and in rarer syn-
hypoplasia, exom-
phalos, severe
dromes such as the femorofacial syndrome.14 However,
hypospadia, FGR the group of diseases that accounts for most cases of
GA 5 gestational age; VSD 5 ventricular septal defect; PKK 5
micrognathia is the chromosomal aberration group. In
polycystic kidneys; TOP 5 termination of pregnancy; FGR 5 fetal fact, micrognathia has been reported in 80% of cases of
growth restriction; NTD 5 neural tube defect; AVSD 5 atrioventricu- trisomy 18 and triploidy at autopsy,1,2 and, conversely,
lar septal defect; CPC 5 choroid plexus cyst; NND 5 neonatal death.
an abnormal karyotype has been found in 66% of
fetuses with micrognathia. This finding was reported in
Discussion 1993 by Nicolaides et al,5 who studied 56 cases of
The growth process of the fetal mandible is impaired in micrognathia detected prenatally, which, to the best of
three groups of diseases: chromosomal aberrations, our knowledge, represent the largest series of fetuses
with micrognathia published to date. In that study, the
authors were able to detect micrognathia by subjective
evaluation of the fetal profile in 25% of fetuses with
trisomy 18 in comparison with the 80% rate reported in

Table 4. Diagnostic Accuracy of the Jaw Index and

Evaluation of the Facial Profile in the Diagnosis of
Micrognathia in the Population of 198 Malformed
Fetuses
Positive Negative
Sensitivity Specificity predictive predictive
Variable (%) (%) value (%) value (%)

Facial profile* 72.7 99.2 80.0 98.8

Jaw index
,24 100.0 98.1 68.7 100.0
,23 100.0 98.1 68.7 100.0
,22 90.1 99.2 83.3 99.6
,21 72.7 100.0 100.0 98.9
Figure 4. Boxplot of the jaw index in normal fetuses and fetuses with * In one case, considered a false negative, it was not possible to
micrognathia. Boxes indicate quartiles (25th and 75th percentiles), and assess the facial profile because of unfavorable fetal lie and oligohy-
vertical bars indicate ranges. dramnios.

VOL. 93, NO. 3, MARCH 1999 Paladini et al Micrognathia in Fetal Life 385
an autopsy series.1 On the basis of this discrepancy, the
authors stated that “at present only the most severe
degrees of micrognathia are amenable to prenatal diag-
nosis.” Since then, several efforts have been made to
improve the ultrasonic diagnosis of fetal micrognathia,
which relied on the subjective evaluation of the profile
on a sagittal midline scan of the fetal face. To render the
diagnosis of micrognathia more objective, the length of
the mandibular bone was measured, and growth
curves, showing a linear relationship with gestational
age and biometric parameters, were developed.3,4 How-
ever, neither Otto et al3 nor Chitty et al4 have further
supported their work with a prospective study. In the
present study, we confirmed the linear relationship
with gestational age and BPD for both the anteroposte-
rior and the laterolateral diameter (Table 1). In addition,
it has been shown in the prospectively studied popula-
tion of 198 malformed fetuses that the measurement of
the jaw index has a greater diagnostic accuracy than the
subjective assessment of the facial profile (Table 4). In
particular, in the diagnosis of micrognathia, a jaw index
of less than 23 has a 100% sensitivity and 98.7% speci-
ficity and a jaw index of less than 21 has a 100% positive
predictive value (Table 3). In previous studies,3,4 the
mandibular growth process was assessed in a single
plane, measuring the length of the bone from the
symphysis mentis to the proximal end of the ramus. In
the present study, the measurement of the two orthog-
onal diameters (laterolateral and anteroposterior) al-
lowed us to analyze separately the two vectors of
mandibular growth (coronal and sagittal). This method
has shown that the lateral growth of the mandible is
relatively regular also in cases of micrognathia and that
the growth process is impaired primarily in the antero-
posterior plane (Figures 2 and 3). We think that this
consideration might explain the relatively high diag-
nostic performance of the anteroposterior mandibular
measurement and, hence, of the jaw index.
References
1. Buyse ML. The birth defects encyclopedia. Boston, Massachusetts:
Blackwell, 1990:1099 –102.
2. Jones KL. Smith’s recognizable patterns of human malformation.
5th ed. London: WB Saunders, 1997:794 –5.
386 Paladini et al Micrognathia in Fetal Life
3. Otto C, Platt LD. The fetal mandible measurement: An objective
determination of fetal jaw size. Ultrasound Obstet Gynecol 1991;
1:12–7.
4. Chitty LS, Campbell S, Altman DG. Measurement of the fetal
mandible—Feasibility and construction of a centile chart. Prenat
Diagn 1993;13:749 –56.
5. Nicolaides KH, Salvesen DR, Snijders RJM, Gosden C. Microgna-
thia fetal facial defects: Associated malformations and chromo-
somal abnormalities. Fetal Diagn Ther 1993;8:1–9.
6. Laitinen SH, Ranta RE. Cephalometric measurements in patients
with Pierre-Robin syndrome and isolated cleft palate. Scand J Plast
Reconstr Surg Hand Surg 1992;26:177– 83.
7. van der Haven I, Mulder JW, van der wal KGH, Hage JJ, de
Lange-de Klerk ESM, Hauman TJ. The jaw index: New guide
defining micrognathia in newborns. Cleft Palate Craniofac J 1997;
34:240 –1.
8. Macpherson TA, Study group for complications of perinatal care
(SGCPC). A model perinatal autopsy protocol. AFIP (Armed
Forces Institute of Pathology), ed. Washington, DC, 1994:234 – 6.
9. Pilu G, Romero R, Reece EA, Bovicelli L, Hobbins JC. The prenatal
diagnosis of Robin anomalad. Am J Obstet Gynecol 1986;154:
630 –2.
10. Franceschetti A, Klein D. Mandibulo-facial dysostosis: New hered-
itary syndrome. Acta Ophthalmol 1949;27:141–224.
11. Reynolds JF. A new autosomal dominant acrofacial dysostosis
syndrome. Am J Med Genet, 1986;25Suppl 2:143–50.
12. Young LW. New syndrome manifested by mandibular hypoplasia,
acro-osteolysis, stiff joints, and cutaneous atrophy (mandibulo-
acral dysplasia) in two unrelated boys. BD:OAS VII(7). New York:
March of Dimes Birth Defects Foundation, 1971:291–7.
13. Turner GM, Twining P. The facial profile in the diagnosis of fetal
abnormalities. Clin Radiol 1993;47:389 –95.
14. Robinow M, Sonek J, Buttino L, Veghte A. Femoral-facial syn-
drome—prenatal diagnosis—autosomal dominant inheritance.
Am J Med Genet 1995;57:397–9.
Address reprint requests to:
Dario Paladini, MD
Via Cimarosa, 69
80127- Naples
Italy
E-mail: paladini@cds.unina.it
Received April 27, 1998.
Received in revised form July 31, 1998.
Accepted August 13, 1998.
Copyright © 1999 by The American College of Obstetricians and
Gynecologists. Published by Elsevier Science Inc.
Obstetrics & Gynecology